|
Utility of blood pressure monitoring outside of the clinic setting |
Appel L J, Robinson K A, Guallar E, Erlinger T, Masood S O, Jehn M, Fleisher L, Powe N R, Bass E B |
|
|
Authors' objectives To investigate the clinical utility of noninvasive ambulatory and self-measured blood-pressure monitoring (ABP and SMBP, respectively) outside the clinical setting for the evaluation and management of adults with elevated BP.
Searching MEDLINE (from 1960 to March 2001), the Cochrane CENTRAL Register (from 1948 to March 2001) and HealthSTAR (from 1975 to March 2001) were searched; the search terms were reported. The January to May 2001 issues of priority journals (identified as those most frequently cited in the database results) were handsearched, and a database of reference material was constructed from relevant guidelines, reviews and discussions with experts. Reference lists and conference proceedings were also consulted. Non-English language articles were excluded. The search strategies were reported in full.
Study selection Study designs of evaluations included in the reviewAll study designs, with the exception of cross-sectional and retrospective studies, were eligible for inclusion. Studies of less than 50 participants (or less than 20 participants where the article addressed reproducibility) were excluded.
Specific interventions included in the reviewStudies of ABP or SMBP interventions were eligible for inclusion. Various measurement devices and strategies were represented in the studies.
Reference standard test against which the new test was comparedThe review did not include any diagnostic test accuracy data to compare the performance of an index test with a reference standard of diagnosis. However, the standard clinic measurement of BP was used as a comparator in the analysis.
Participants included in the reviewWith the exception of pregnant women and children under 20 years of age, the review was not limited by age, gender, or other patient characteristics. A variety of participants were included in the review, e.g. those with treated and untreated hypertension, non-hypertensives and diabetics.
Outcomes assessed in the reviewStudies that provided original data (to address specific research questions) on BP measurement and the longitudinal assessment of clinical outcomes were eligible for inclusion. The outcomes included target organ damage (left ventricular index mass or proteinuria/albuminuria), symptoms, white coat hypertension (WCH), morbidity and mortality. WCH was described as an elevation of BP in the clinic setting, which is not sustained outside of this setting. Clinical trials that did not have longitudinal analyses of outcomes other than BP were excluded.
How were decisions on the relevance of primary studies made?The review process was conducted in three stages. In the first stage, two reviewers scanned titles and abstracts. In the event that eligibility could not be agreed, the abstract was deemed eligible for full article review. This process was repeated at a second stage, where more stringent exclusion criteria were applied. The large number of full articles deemed eligible for inclusion meant that further exclusion criteria were subsequently applied at the (third) article review stage. Details of the exclusion criteria (applicable to all studies, and specifically to each research question) were given. The reviewers were not blinded to the study author, institution or journal.
Assessment of study quality The authors devised a quality assessment checklist from previous projects, discussions with experts and through pilot testing. Different assessment forms were developed to assess prospective studies, studies of reproducibility and trials examining treatment effects. The quality assessment forms used were reproduced in the appendices of the report. A primary reviewer completed the quality assessment, while a secondary reviewer checked the extracted data for completeness and accuracy.
Data extraction A serial article review process was employed. As for the validity assessment, a primary reviewer completed the data extraction and a secondary reviewer checked the data. Various data were collected, according to each research question. For example: data related to different methods of BP measurement; mean differences with standard deviations; levels of observer or patient training; definitions of hypertension; target organ damage; and morbidity and mortality. Where necessary, the data were transformed to meet the requirements of the review, e.g. for the calculation of relative risks from incidence rates. Not all relevant information was available from all included studies.
Methods of synthesis How were the studies combined?The studies were combined in narrative according to specific research questions.
Question 1: comparison of clinic BP, SMBP and ABP readings.
What is the distribution of the BP differences between clinic, ambulatory and SMBP readings?
What is the prevalence of WCH as defined by SMBP and by ABP measurement?
What is the reproducibility of differences in readings and WCH?
Question 2: the relationship of SMBP/ABP levels and WCH based on SMBP/ABP with target organ damage and clinical outcomes.
Is SMBP/ABP more or less strongly associated with BP-related target organ damage than clinic BP measurements?
Does SMBP/ABP predict subsequent clinical outcomes?
What is the incremental gain in the prediction of clinical outcomes from the use of self-measurement or ambulatory devices beyond prediction from clinic BP alone?
What is the effect of treatment guided by SMBP/ABP in comparison with treatment guided by clinic BP?
Question 3: subgroup analysis.
Does the evidence for the above questions vary according to a patient's age, gender, income level, race or ethnicity, and clinical subgroup?
How were differences between studies investigated?Differences between the studies (e.g. measurement device, location, sample size and other study characteristics) were discussed narratively.
Results of the review Eighty-nine studies were included in the review. There was no clear summary of how many studies of each design were included, or how many participants were included overall and in each study type.
Question 1: comparison of clinic BP, SMBP and ABP readings.
Eighteen studies investigated the distribution of readings between the three measurement methods. For both systolic and diastolic BP measurements, readings taken in the clinic setting were significantly higher than those taken from self-measurement or ambulant methods (P<0.05). Readings were similar when SMBP and ABP were compared, but this occurred in only 3 studies. It was not possible to determine whether BP differences were reproducible. From 16 studies that addressed the prevalence of WCH (various definitions) based on ABP, the largest of these (using the greatest number of measurements) revealed a rate of 20% amongst hypertensive patients. An insufficient number of studies meant that neither the reproducibility of the latter, nor the prevalence of WCH based on SMBP, could be determined.
Question 2: the relationship of SMBP/ABP levels and WCH based on SMBP/ABP with target organ damage and clinical outcomes.
Results from 25 studies revealed that left ventricular mass and albuminuria were positively correlated with ABP. There was not enough evidence to determine the strength of associations based on SMBP. In 10 prospective studies, at least one dimension of ABP successfully predicted a clinical event (usually cardiovascular disease). WCH was associated with a reduced risk of the event (relative to sustained hypertension) (2 studies) and a non-dipping/inverse measurement pattern suggested an increased risk (4 studies). Again, there was inadequate evidence for the relationship between SMBP and WCH and subsequent clinical outcome. Only 2 of 9 prospective studies addressed incremental gain for ABP and only 1 study for SMBP over clinic BP measurement in the prediction of risk. All studies were hindered by poor or uncertain measurement quality, and conclusions could not be reliably drawn. Twelve trials addressed the effect of treatment guided by SMBP and 2 trials assessed treatment guided by ABP. Although there was some suggestion that treatment guided by SMBP could lead to reduced BP, the evidence was either inconsistent or insufficient for reliable conclusions to be drawn.
Question 3: subgroup analysis.
Most studies included both men and women. It was noted that African-American participants were under-represented. There was no stratification of the results by race or gender, except in one analysis where a higher prevalence of WCH was noted in women than in men.
Authors' conclusions ABP levels and patterns were associated with BP-related target organ damage and an increased risk of subsequent cardiovascular disease events. There were too few studies to determine the evidence for these relationships in the case of SMBP. Comparisons of risk prediction based on ABP or SMBP and clinical BP were also precluded by insufficient data. ABP is a potentially useful prognostic tool. The extent to which risks associated with WCH are low enough to preclude drug therapy in hypertensive patients was unclear.
CRD commentary The questions and the inclusion criteria were clear for this substantial review. The search strategy was comprehensive in its consultation with a number of databases and additional material. However, the restriction to English language articles may mean that important studies might have been missed, and there was no evidence that publication bias was assessed. Other potential sources of bias were noted in the review process; specifically the apparent lack of independent (and non-blinded) assessment at the study selection, data extraction and validity assessment stages. Although the study details were reported extensively and the results were clearly synthesised according to research question, there was no clear summary of study quality. Together with the fact that many of the findings were based on studies with substantial heterogeneity and inconsistent levels of reporting, the reliability of the conclusions is unclear. However, the authors substantiated this as a likely reflection of practice and offered a comprehensive list of recommendations for future research.
Implications of the review for practice and research Practice: The authors stated that this review will aid policy makers in their evaluation of alternative strategies to measure BP.
Research: The authors stated a number of recommendations for future research. These included more prospective observation studies to look at (for example) the risks and reproducibility of WCH and incremental gains of measurements taken outside the clinic setting. They also proposed more clinical trials to test contemporary SMBP technology; decision analyses to take account of economic factors; more evidence on the accuracy and performance of clinic-based BP measurement strategies; and feasibility studies to assess ABP and SMBP in routine use.
Funding Agency for Healthcare Research and Quality, contract number 290-97-006.
Bibliographic details Appel L J, Robinson K A, Guallar E, Erlinger T, Masood S O, Jehn M, Fleisher L, Powe N R, Bass E B. Utility of blood pressure monitoring outside of the clinic setting. Rockville, MD, USA: Agency for Healthcare Research and Quality. Evidence Report/Technology Assessment; 63. 2002 Other publications of related interest This additional published commentary may also be of interest. Powers DV. Review: ambulatory blood pressure monitoring predicts clinical outcomes. Evid Based Med 2003;8:120.
Indexing Status Subject indexing assigned by CRD MeSH Blood Pressure; Blood Pressure Monitoring, Ambulatory AccessionNumber 12003008164 Date bibliographic record published 31/10/2005 Date abstract record published 31/10/2005 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|