Application of ambulatory blood pressure monitoring in public primary care clinics in Hong Kong: what do primary care doctors need to know?

Kwai-sheung Wong 黃桂嫦，Ka-ming Ho 何家銘，Yim-chu Li 李艷珠，Catherine XR Chen 陳曉瑞

HK Pract 2022;44:81-89

Summary

Objective: To delineate the indications for ordering ambulatory blood pressure monitoring (ABPM) in public primary care setting and to explore patient characteristics and results of the ABPM.
Design: Cross-sectional descriptive study.
Subjects: All patients who have had ABPM performed from 1/12/2016 to 30/11/2017 in five General Outpatient Clinics.
Main outcome measures: The indications for performing the ABPM, demographics of patients undergoing ABPM and the results of ABPM were studied.
Results: There were 323 patients with ABPM done within the study period with valid results. 64% were female and 36% were male, and the average age was 64 ± 12 (19 to 94 years old). 67 (21%) with diabetes mellitus and 53 (16%) with impaired glucose tolerance or impaired fasting glucose.

For the indications for ABPM, 150 (46%) were for establishing diagnosis of hypertension (HT) and 173 (54%) were for monitoring of blood pressure (BP) control among hypertensive patients.

Among the diagnosis of HT group, 96 (64%) were confirmed with the diagnosis of HT, 42 (28%) were found to have white-coat HT only.

For the monitoring of BP control group among HT patients, 98 (57%) were noted to have suboptimal BP control and 67 (39%) were found to have white-coat effect.

Among the 65 patients whose ABPM had been ordered although their clinic BPs being normal, 18 (28%) were diagnosed to have masked HT and 19 (29%) were diagnosed to have masked uncontrolled HT.

Keywords: Ambulatory blood pressure monitoring, hypertension, primary health care

摘要

目的 ： 規劃在公共基層醫療環境中24小時動態血壓監測 (ABPM)的用途，並探索24小時動態血壓監測病人的特徵 和結果。
設計 ： 橫斷面式描述性研究。
對象 ： 從2016年12月1日至2017年11月30日期間，所有在 五個普通科門診診所進行了24小時動態血壓監察的病人。
主要結果測量 ： 研究了進行廿四小時動態血壓監察 的原因、病人的人口統計學特徵及結果。
結果 ： 在研究期間，有323位進行了廿四小時動態血 壓監察的病人的結果為有效。64%為女性，36%為男 性，平均年齡為64±12歲(19至94歲。67人(21%)患有 糖尿病，53人(16%)患上葡萄糖耐性障礙或空腹血糖 偏高。

至於進行二十四小時動態血壓監察的原因，150名 病人(46%)用於確定高血壓(HT)的診斷，173名病人 (54%)用於監測高血壓患者的血壓(BP)控制。 關於血壓高的的診斷中，96名病人(64%)被診斷為血 壓高，42(28%)被發現只有白袍高血壓。

關於血壓高病人的控制監察方面，98名病人(57%)的 血壓控制不理想，而67名病人(39%)被發現具有白袍 效應。 儘管診所血壓正常，但仍然進行二十四小時動態血 壓監察的65名患者中，18名(28%)被診斷出患有隱性 高血壓，19名(29%)被診斷出患有不受控制的隱性高血壓。

關鍵詞 ： 廿四小時動態血壓監測、高血壓、基層醫療

Introduction

Hypertension (HT) is a common chronic disease affecting more than 20% of adults globally¹ and more than 27% of Hong Kong (HK) people aged 15-84.² It is also one of the most common clinical conditions encountered in primary care. Numerous studies have shown that HT is one of the major cardiovascular risk factors.^3,4 For every 10 mmHg reduction in systolic blood pressure (BP), the risk of major cardiovascular disease events reduced significantly by 20%, leading to a significant 13% reduction in all-cause mortality.⁵ Therefore, proper diagnosis and management of HT is an important mission for primary care doctors.

Traditionally, HT is diagnosed based on office or clinic BP readings. Recent studies have revealed that ambulatory blood pressure monitoring (ABPM) could provide better prognostic information about cardiovascular disease than office BP measurement.^6,7 As ABPM can provide multiple BP measurements away from the medical environment, it gives a clearer picture of the patient’s BP profile over a 24-hour period including both awake and sleep time. Such valuable information cannot be offered by office BP and home BP measurements. This allows the identification of individuals white-coat HT (WC HT), HT with white-coat effect (HT with WC effect), masked HT and masked uncontrolled HT. Therefore, some prestigious international guidelines on management of HT have recommended the use of ABPM in establishing the diagnosis of HT or monitoring BP control in existing hypertensive patients.^8,9.10 Indeed, ABPM has been shown to be cost-effective both in specialist services and in primary care, with large amount of resources being saved due to more rational drug prescribing and better BP control.¹¹ However, despite all of this evidence, the application of ABPM in primary care is still suboptimal. For example, a review of ABPM use in 11 Asian regions including China, Japan, Korea and Singapore published in the year 2019 noted that ABPM was seldom available in primary care clinics.¹² It is therefore not surprising to find that ABPM has been underused in the primary care setting.

ABPM services has been introduced to public General Outpatient Clinics (GOPCs) of Kowloon Central Cluster (KCC) under the Hospital Authority of Hong Kong (HAHK) since 2015. It assists in the diagnosis of HT in suspected hypertensive patients and monitoring of BP control in hypertensive patients. Until this moment, the service utilisation and the results of the ABPM performed in local primary care clinics had not been widely studied. To fill in this knowledge gap, this study aims to delineate the indications for ordering ABPM in the public primary care setting and to explore the patient characteristics and the results of the ABPM. We believe that the results of our study will provide important background information for the use of ABPM in public primary care clinics and enlighten family physicians on its proper application and interpretation in HT management.

Methods
Study Design:

Cross-sectional descriptive study carried out at public primary care clinics

All patients with ABPM done within the period from 1/12/2016 to 30/11/2017 in five GOPCs of KCC under the Hospital Authority of HK were included. It was estimated that around 14000 hypertensive patients attended these clinics in the period. ABPM would be ordered in GOPC directly by the attending doctors if indicated. Patients who were intolerant to ABPM or whose ABPM data was invalid were excluded.

ABPM measures blood pressure monitoring patients’ BP during a 24-hour period. “ABPM awake” and “ABPM sleep” were defined according to the patients’ schedule. All patients who have undergone ABPM were asked about the awake and sleep time and these data were entered into an ABPM software (cardioversions 1.16.6). They were advised not to do vigorous exercise during ABPM. Awake BP readings were measured in 30-minutes intervals. Sleep readings were measured in 45-60-minutes intervals, depending on the total number of sleeping hours, with a minimum of 7 measurements during sleep. The minimum awake measurements should not be less than 20. The ABPM would be considered as invalid if the overall successful measures were less than 70% of all readings.¹³

   Indications for ambulatory blood pressure monitoring

    ABPM is usually indicated in the following scenarios in our clinics based on the recommendations from the NICE guidelines⁸:

1. To establish the diagnosis of HT, especially if patients do not have out-of-office BP measurements.
2. To monitor the BP control among hypertensive patients, such as those with unusual discrepancy between their clinic BP measurement and home BP measurement,when there are symptoms suggestive of hypotension, or when evaluating drug resistant HT.

   Sample size calculation

   The prevalence of white-coat HT in the community was reported to be 20 – 25% in the literature.¹⁴ Assuming the proportion of white-coat HT is 25% in primary care clinics, margin of error being 5% and confidence level of 95%, by using the formula:

(https://www.calculator.net/sample-size-calculator.html; z is the z score, ε is the margin of error and p̂ is the population proportion), the required sample size is 289. To allow room for case exclusion (~15%), 340 samples were included in the study.

Determination of variables

All relevant clinical data were retrieved from the Clinical Management System (CMS) of HAHK. The demographic data of the patients including age, gender and comorbidities with diabetes (DM) or pre-diabetes including impaired fasting glucose and impaired glucose tolerance were retrieved. The patients’ BP reading on the date of ABPM ordered, the indications of the ABPM and the final diagnosis after the ABPM was done were also collected. According to the latest NICE guideline⁸ and American Diabetes Association guideline in 2020¹⁵,cinic BP is defined as normal if the reading is <140/90mmHg either in the presence or absence of DM.

Statistical analysis

Statistical analysis was performed using the Statistical Package for the Social Sciences (Windows version 26.0; SPSS Inc, Chicago [IL], US). All continuous variables are expressed as mean ± standard deviation. Categorical data are reported as number and percentages. Chi-square test was used for comparing the categorical variables. A P-value of less than 0.05 is considered as statistically significant.

Results

349 patients had ABPM performed during the one-year study period, among which 26 patients (7%) were excluded from data analysis as their ABPM results were invalid. Table 1 summarised the characteristics of the remaining 323 patients (93%) included in the data analysis. Among them, 208 (64%) were female and 115 (36%) were male. They were aged 19 - 94 years old, with the mean age being 64 ± 12. Their mean clinic systolic BP was 151 ± 18 mmHg and mean clinic diastolic BP was 81 ± 12 mmHg. 65 patients (20%) were found to have normal clinic BP readings at the time when ABPM was ordered. Furthermore, 67 (21%) were found to have concomitant DM and 53 (16%) with pre-diabetes including impaired glucose tolerance or impaired fasting glucose. The average waiting time for the ABPM service was 26 days.

   As to the indications for ABPM, 150 (46%) were for establishing a diagnosis of HT and 173 (54%) were for monitoring of BP control among confirmed HT patients. For those whose ABPM were ordered for diagnosing of HT, 96 (64%) were confirmed to have a diagnosis of HT, 42 (28%) were found to have white-coat HT, i.e. elevated clinic BP with normal ABPM and 12 (8%) were found to be totally normotensive (Figure 1). The age and gender distribution among the subgroups were comparable (Table 2).

   For those whose ABPM were indicated for monitoring of BP control among hypertensive patients, 98 (57%) were noted to have suboptimal BP control and 67 (39%) were hypertensive patients with white-coat effect, i.e. elevated clinic BP but normal ABPM in treated HT individuals (Figure 2). Again, the subgroup analysis as shown in Table 3 revealed that their age and gender condition were comparable.

   Sixty-five patients had ABPM ordered although their clinic BPs were normal (<140/90mmHg). Among them, 37 (57%) were found to have elevated ABPM, with 18 (28%) having masked HT and 19 (29%) having masked uncontrolled HT (Figure 3).

Discussion

Although wide evidence from literature has confirmed the benefit of ABPM in establishing the diagnosis of HT and the management of different types of HT, its application in primary care is still far from satisfactory. For example, a cross-sectional survey of more than 400 primary care doctors in HK in 2019 found that there was a significant underutilisation of ABPM, with only 1.6% doctors using this method to diagnose HT and 4.1% using this for management of HT.¹⁶ In our study, due to limited resource in the public primary care system, only 349 cases had ABPM done during the one-year study period and this figure was definitely suboptimal in view of the large numbers of HT patients management in the GOPCs. We hope our study would help delineate the role of ABPM by reviewing the application of ABPM in local primary care clinics and therefore advocate its use among all primary care doctors in HK.

Our study reveals that nearly one third of untreated individuals were noted to have white-coat HT and more than one third of hypertensive patients were found to have white-coat effect. These findings are consistent with a local study done in 2007 showing that 28.2% of all HT patients had white-coat HT in primary care.¹⁷ Similarly, in another local study done by Chiang et al in 2013, 30% of treated hypertensive patients in a public GOPC were found to have white-coat effect.¹⁸ Currently, there is insufficient evidence from randomised controlled trials to determine whether white-coat HT warrants treatment.¹⁹ Dietary advice and lifestyle modifications were recommended as the mainstay of treatment for white-coat HT.²⁰ With the help of ABPM, over-diagnosis of HT could be avoided in nearly 30% of cases and therefore prevent the wrong labelling of “HT” among this group. Similarly, the use of ABPM helps ease the worry of those HT patients with W-C effect and subsequently avoid the unnecessary medication augmentation. This is particularly important as otherwise the patients will be at risk of hypotension due to overtreatment and be susceptible to the potential side-effects from drug treatment. All these findings have corroborated the importance of doing ABPM before establishing the diagnosis of HT or when white-coat effect is suspected among treated HT patients.

Importantly, more than half of the patients were found to have suboptimal BP control in the group of hypertensive patients whose ABPM were ordered for monitoring of BP control. This result is consistent with findings from another study showing that more than 40% of patients labelled as having “HT with white-coat effect” were actually having suboptimal BP control in the local primary care clinics.²¹ This finding should alert primary care doctors to the importance of conducting ABPM for confirming the BP control status before the clinical decision of “satisfactory control” is made liberally. Although regular home BP monitoring (HBPM) could help confirm the white-coat effect among hypertensive patients too, self-reporting bias, patients with impaired cognition and the lack of monitoring during the sleep time by typical HBPM machine may prompt the use of ABPM rather than HBPM.²⁰ In view of the fact that poorly controlled BP will significantly increase hypertensive patient’s cardiovascular risks, family physicians should carefully review their BP control status during every follow-up visit and refer the case to have ABPM if in doubt.

It is alarming to note that quite a number of patients referred to ABPM with normal clinic BP were found to have either masked HT (28%) or masked uncontrolled HT (29%). A local study by CM Ng et al involving type 2 diabetic patients who were not on any antihypertensive in a tertiary clinic found that the prevalence of masked HT was 18% by ABPM.²² The high percentage of masked HT or masked uncontrolled HT of our study might be explained by several reasons. Firstly, when ABPM was referred for patients whose clinic BP was normal, their home BP readings were usually elevated. The discrepancy between the clinic BP and home BP readings would prompt the attending doctor to further assess their BP control by ABPM. As the BP machines in the study clinics are under regular maintenance and validation service provided by the Hospital Authority, we assume the clinic BP machines and measurements were accurate in our study. Secondly, ABPM would be ordered for patients having early development of cardiovascular complications or chronic kidney disease (CKD) with seemingly well controlled HT. This is a sensible choice as many such patients were later confirmed to have masked HT or masked uncontrolled HT by ABPM and therefore warranting closer BP monitoring and stricter control. Such people usually have more metabolic risk factors and asymptomatic organ damage than those who are truly normotensive.²³ This is particularly important in view of the large amount of evidence showing that masked HT significantly increases the cardiovascular morbidity.^24,25 The long-term risk of developing sustained HT, DM, or left ventricular hypertrophy in masked hypertensive patient is 2 to 3 times greater than that of individuals with normal in- and out-of-office BP.²⁶ With all these findings, we would like to appeal to all primary care doctors to enhance the use of out-ofoffice BP measurements including ABPM when there is reasonable doubt on the BP control although the clinic BP is normal. An accurate and timely assessment of BP control could prevent or delay the development of cardiovascular complications.

Limitations of the study

There are some limitations in this study. First, the study was carried out in one single cluster of HAHK, therefore these results from the public primary health care sector might not be applicable to the private sector or secondary care. In addition, only 323 cases who fulfilled the ABPM referral criteria and with valid ABPM results were included in the data analysis, therefore selection bias might exist. Nevertheless, as the data included all patients who had done ABPM from five primary care clinics in HAHK during the one-year study period, these data may give a realistic representation of ABPM service performed in the public primary care settings and had provided important background information for future service enhancement. Secondly, apart from DM and pre-diabetes, data on other comorbidities such as ischemic heart disease, stroke or CKD etc. had not been collected. It is therefore hard to determine the correlations between the ABPM findings with the cardiovascular or renal complications. Future studies with more comprehensive data collection will be needed to further explore this important aspect. Thirdly, due to feasibility reasons, not all participants were able to do the ABPM on the same day and their average waiting time was 26 days. Patient’s clinic BP on the date when ABPM was ordered might be substantially different from the date when ABPM was performed. Having said so, since most cases who attended the GOPCs were comparatively stable cases and therefore the four weeks’ time interval is a reasonable timeframe where patients’ clinical condition would unlikely to have dramatically changed. Lastly, because of the crosssectional nature of the study, we were unable to adjust for potential unmeasured confounders and therefore no temporal or causal relationship could be established.

   Conclusion

   Our study successfully delineated the indications for ordering ABPM in the public primary care setting, explored the patient characteristics and the results of the ABPM. ABPM was used for both establishing the diagnosis of HT and monitoring of BP in hypertensive patients. More than one third of primary care patients undergoing ABPM were found to have either white-coat HT or HT with white-coat effect. The use of ABPM might help avoid unnecessary initiation or escalation of antihypertensive drugs among these groups of patients. In addition, more than half of treated HT patients undergoing ABPM were found to have suboptimal BP control, highlighting the important role of ABPM in the monitoring of BP control among HT patients. Furthermore, more than half of patients referred to ABPM with normal clinic BP were found to have masked HT or masked uncontrolled HT. Further study is needed to explore the risk factors contributing to the development of masked HT and to prevent or delay the development of cardiovascular complications among this group of patients.

Acknowledgements

References