IRCCS Istituto Auxologico Italiano.
Department of Statistics and Quantitative Methods.
J Hypertens. 2019 Mar;37(3):467-475. doi: 10.1097/HJH.0000000000001914.
Office and ambulatory blood pressure (BP) measurements are the main techniques to detect the effects of antihypertensive treatments in clinical trials, but the treatment-induced changes in these BP values can differ markedly. We performed a meta-analysis of clinical trials to quantify these differences and identified some of the associated factors.
We conducted a MEDLINE search for randomized clinical trials (RCTs) on hypertensive patients treated with at least one antihypertensive drug that reported changes in both office and 24-h BP. Random-effects models were fitted to estimate the summary of the difference between the changes as quantified by either technique. The I and Cochrane's Q statistics were calculated to evaluate the heterogeneity between studies.
A total of 52 studies were included in our meta-analysis with about 9500 patients. The summary estimate Δ of SBP and DBP was -6.5 (95% confidence interval: -7.5 to -5.6) and -3.3 (95% confidence interval, -3.9 to -2.7), respectively. The difference was independent on the treatment duration and use of mono or combination treatment but for SBP it varied with the different treatment types in monotherapy, and it was greater in relation to baseline office BP and age. Allowing for the placebo effect also reduced the difference. Absolute on treatment BP values were at target for either pressure and the rate of controlled hypertensive individuals was similar (around one-third) for either measuring approach.
Our meta-analysis confirms that overall treatment-induced reduction is markedly greater for office BP than for 24-h BP, but it also shows that the quantitative relationship between these two measuring approaches varies with demographic, clinical and therapeutic conditions as well as in relation to placebo correction.
诊室血压（BP）和动态血压（ABP）测量是检测降压治疗临床试验效果的主要技术，但这些 BP 值的治疗诱导变化可能有显著差异。我们对临床试验进行了荟萃分析，以量化这些差异，并确定了一些相关因素。
我们对使用至少一种降压药物治疗的高血压患者的随机临床试验（RCT）进行了 MEDLINE 搜索，这些 RCT 报告了诊室和 24 小时 BP 的变化。使用随机效应模型拟合估计汇总差异，以定量两种技术之间的变化。计算 I ²和 Cochrane's Q 统计量以评估研究之间的异质性。
共有 52 项研究纳入我们的荟萃分析，涉及约 9500 名患者。SBP 和 DBP 的汇总估计值Δ分别为-6.5（95%置信区间：-7.5 至-5.6）和-3.3（95%置信区间：-3.9 至-2.7）。差异独立于治疗持续时间和单药或联合治疗的使用，但在单药治疗中，不同治疗类型的 SBP 差异存在差异，且与基线诊室 BP 和年龄有关。考虑安慰剂效应也会降低差异。两种测量方法的治疗后 BP 值均达到目标，两种测量方法的血压控制率相似（约为三分之一）。
我们的荟萃分析证实，与 24 小时 BP 相比，整体治疗诱导的降低在诊室 BP 中更为显著，但也表明这两种测量方法之间的定量关系随人口统计学、临床和治疗条件以及与安慰剂校正有关而变化。
