The current study found a great discrepancy between BP measuring methods recommended by international guidelines and those used in routine clinical practice by general practitioners in HK, who treat most people with hypertension. Majority of general practitioners used manual office BP-measuring methods, particularly using mercury or aneroid devices, for routine screening (63.1%), diagnosis (56.4%), and management (72.4%) of HT. Furthermore, while using office BP, many doctors (26.7%) reportedly obtained only one BP reading for diagnosis and/or treatment. Similarly, 60% doctors took <12 HBPM readings for diagnosis or management of HT. On the other hand, ABPM was prominently underutilised, with only ~1% of doctors using this method to diagnose HT. The other out-of-office BP measurement method, HBPM, was used by less than one-fifth of doctors for diagnosis (22.2%), though around half of the respondents (56.8%) used HBPM to guide management of HT.
Despite manual office BP measurement being most commonly used by our respondents, it is the least accurate of all methods and therefore is least able to predict cardiovascular events. Further, it is prone to assessment errors due to factors such as inadequate rest period prior to BP measurement, carrying on a conversation during BP measurement, obtaining inadequate numbers of BP readings, rounding-errors when recording BP value, and excessively rapid manual deflation of the BP cuff (22).
The over-dependence on manual office BP and the underutilisation of out-of-office BP measurements mean that a substantial proportion of patients are over-treated or undertreated if they have white-coat hypertension or masked hypertension, respectively (4). Researchers have shown that ≤30% patients could receive unnecessary treatment, if hypertension is solely diagnosed based on office BP values, as they could simply have white-coat hypertension (4). A few studies have suggested the underlying reasons behind this preference for BP measurement methods. An Australian primary care study found that general practitioners were uncertain about the best way to measure BP, were unsure about the cut-off values for out-of-office BP estimation, did not have enough time to discuss techniques or results of ABPM or HBPM with the patients, and did not have the resources to prescribe or provide HBPM/ABPM (23); these general practitioners suggested that a dedicated primary care guideline was needed (23). Similarly, doctors are often concerned about whether the patient could master HBPM techniques and follow the strict measurement protocols required, to obtain accurate readings (24). ABPM may also be perceived by doctors as inaccessible, expensive, inaccurate, and poorly tolerated (24). Similar research in Chinese and Asian populations, especially in Hong Kong, is lacking. Although, several studies have described possible barriers to using out-of-office BP measurement methods, more studies are needed to understand how to implement accurate BP assessment techniques (e.g. ABPM/AOBP) in primary care. The current study also found that family medicine specialists were more likely to use AOBP in clinics and also to obtain enough office BP readings for diagnosis and management of HT, suggesting that targeted medical education can modify BP-measurement behaviour.
Only a few similar studies exist in the available literature. An online survey, which was conducted by The College of Family Physicians of Canada and involved 774 family physicians (response rate, 16.2%), found that AOBP was a common method to screen (42.9%), diagnose (31.1%), and manage (59.2%) HT, while ABPM was reported as the primary diagnostic tool for HT by 14.4% of the respondents (18). A recent study in the United Kingdom (UK) surveying 489 patients, who self-reported to have BP measured during their last clinic visit, found that only one BP reading was obtained in 286 (59.6%) patients (25). Considering the findings of the UK study, around 70% of doctors in our study obtained more than 1 clinic reading when manual BP measurement was used, which was higher than the UK study (25). However, the authors of the UK study commented that their local guideline suggested duplicate office BP measurements only in patients with high BP readings, and this may explain the low duplicate office BP measurement rate (25). Moreover, considering the findings of the Canadian study, significantly fewer doctors in HK used AOBP and ABPM for diagnosis and management of HT (18). The reasons for the choice of BP measurement method cannot be delineated from the current study, and relevant studies on the topic are lacking in the Asian/Chinese population. However, we hypothesised that this may be due to a lack of relevant guidelines in HK. Although, the HK primary care guidelines have discussed the technique of HBPM at length, both ABPM and AOBP are not mentioned; the reasons behind omitting ABPM and AOBP were not discussed in the guideline (17). In this study, despite underuse of out-of-office BP measurements to diagnose HT, more than half of the doctors (56.8%) incorporated HBPM readings to manage HT, suggesting the effect of the local guideline and feasibility of using HBPM in the local population. Secondly, most of our study participants were old and worked in the private sector. These doctors may have used the manual technique over many years and considered it a way to build rapport with patients and demonstrate their expertise (26). Furthermore, patients may be used to, and will expect manual BP measurements, due to its widespread local use (26). Finally, doctors in the private sector are able to conduct patient-consultation for longer periods and can afford to spend time taking systematic manual measurements. These factors may at least partially explain the significant underutilisation of ABPM and AOBP in primary care in HK. A detailed qualitative study in HK, is needed to further ascertain our results and hypotheses.
Despite this being one of the first few studies to describe doctors’ preferences with respect to BP measurement methods, the study has certain limitations. Although, our response rate was low at 25.6%, this is comparable to, or even higher than those observed in previous similar surveys and e-mail studies involving doctors (18,19). The reason behind such low response rate could not be delineated by the current data but may have been due to reasons such as: (i) provision of incorrect mailing addresses, which meant that the questionnaire did not reach the doctors; (ii) the doctors were too busy to respond; (iii) doctors who did not usually treat patients with hypertension might have found the study irrelevant to them; and (iv) doctors who were unsure about their technique of measuring blood pressure correctly and might not have wanted to respond to the questionnaire. Furthermore, the primary care directory did not contain detailed demographic data and it was unclear if the demographics of our participants were similar to those of the non-respondents. Secondly, our results might have overestimated the number of doctors who used AOBP or out-of-office BP measurements because (i) the respondents were more likely to be interested and knowledgeable in BP measurements than non-respondents; and (ii) to ensure social desirability, respondents might have answered the questionnaire according to the international guidelines rather than describing techniques used in their actual clinical practices. However, these factors make our conclusion of underutilisation of out-of-office BP and AOBP measurements, even more robust. Thirdly, most of our respondents worked in the private sector and the applicability of our results to the doctors working in the public sector is unknown. Doctors working in the public sector might not have responded to the questionnaire because, they often do not have the freedom to choose a particular BP measurement method, which is determined by their respective departmental authority. The use of out-of-office BP measurement in the public sector is further limited by availability of resources and it is predicted that the use of out-of-office BP readings in these clinics is even less common. For example, the current public primary care department guideline only recommends ABPM (free of charge) in patients with suspected white-coat hypertension and the waiting time for these patients is currently more than 1 year. Similarly, patients seen in these government-funded clinics usually lack resources or higher education and thus may not be able to afford self-financed HBPM, although HBPM machines can be easily bought in HK. Furthermore, primary care doctors working in public sector have very short consultation time-intervals (often only 2–5 minutes/patient) and limited space (i.e. a quiet room where the patients could stay for 5 minutes for AOBP measurements), which are common barriers to AOBP and out-of-office BP measurements (27). On the contrary, doctors working in the private sector can choose from a range of available BP measurement methods, have the time to counsel patients, and in turn, their patients might be more willing to bear the expense and effort required for ABPM/HBPM. In HK private sector, the cost of out-of-office BP measurements is often paid by patients and there is currently no reimbursement from the insurance companies or government. Lastly, we have only included doctors registered in the primary care directory and it is not known how representative this sample is to the whole population of primary care doctors in HK.