Hypertension prevalence as a function of different guidelines, India

Abstract Objective To determine the effect of different hypertension management guidelines and of basing diagnosis on a single reading of blood pressure on the hypertension prevalence in the Indian population. Methods We performed a secondary analysis of data acquired as part of the Fourth national family health survey, 2015 to 2016, over all districts in India. We calculated the proportion of the population within three different age groups (18 to 34, 35 to 49 and 18 to 49 years of age) with raised blood pressure according to six different guidelines, and how prevalence changed if diagnoses were based on a single blood pressure measurement. Findings We observed that the Government of India and the American College of Cardiology/American Heart Association guidelines consistently yielded the lowest and highest prevalence of raised blood pressure; in the combined age group, we calculated the proportion of the population categorized as having raised blood pressure as 7.5% (95% confidence interval (CI): 7.4 to 7.7) and 40.1% (95% CI: 39.7 to 40.7), respectively. When basing diagnosis on a single reading of blood pressure only, a total of 56 million individuals would be erroneously categorized as hypertensive following the Government of India guidelines. We also showed that prevalence of hypertension in India varies with guidelines adhered to; in the combined age group, the national hypertension prevalence was three times higher when following the American College of Cardiology/American Heart Association compared with the Government of India guidelines. Conclusion To optimize current clinical practice, health-care providers need to follow universally agreed, evidence-based methods of diagnosing hypertension.


Introduction
With developments in technology and the expansion of treatment options and modalities, the field of clinical care guidelines is constantly evolving. Although clinical care guidelines are only recommendations, the decision to follow a specific set of guidelines by a health-care provider should be based on the local context of need, availability and affordability, especially in low-and middle-income countries. 1 The availability of different guidelines, with inconsistencies in recommendations of when medical treatment should be initiated, can cause friction between health-care provider and patient. 2,3 Blood pressure measurement is one of the most common non-invasive clinical practice tools used to assess the cardiovascular status of an individual and predict the likelihood of future cardiovascular events. An individual's blood pressure can change quickly and regularly, and is influenced by respiration, temperature, bladder distension, pain levels, emotion, diet, time since last exercise and whether alcohol has recently been consumed. 4 A simple method of obtaining an accurate blood pressure measurement is to take repeated readings over multiple visits. Compared with diagnoses of hypertension based on a single measurement of blood pressure, studies have demonstrated as much as 12% reduction in the prevalence of hypertension if repeated readings over multiple visits are considered. 5,6 By considering multiple readings, up to 35% of patients were reclassified within a lower category of blood pressure. 5,6 However, even after obtaining a more accurate blood pressure measurement, clinical care guidelines differ with respect to the precise blood pressure at which a patient is diagnosed as hypertensive and begins treatment. [7][8][9][10][11][12] This lack of uniformity between the various available guidelines diminishes the value of measuring blood pressure.
Over the past 25 years, the availability of health-care services have increased in India, and the country has adopted a universal health coverage programme. 13 Although the availability of health-care services has risen, the quality of treatment received from different health-care providers is not consistent. 1,[14][15][16] The causes of this inconsistency in quality across India include variations in clinical practice, poor diagnostic facilities, a lack of expertise, unnecessary use of medicines (e.g. antibiotics, analgesics and steroids) and substandard treatment. 1 India is currently experiencing an increase in the prevalence of noncommunicable diseases, such as hypertension and diabetes, and the accompanying premature mortality. [17][18][19][20][21] Inconsistent guidelines introduce uncertainty in the accuracy of hypertension diagnoses and increase the likelihood of poor health outcomes. 22 Poor healthcare literacy, high self-medication rate, poor blood pressure control and inconsistent hypertension management guidelines intensify the problem in India. 23 Here we have analysed the impact of inconsistent practices on the calculated prevalence of hypertension. We have focused on the particular blood pressure measurements at which hypertension is diagnosed and whether a single reading or the recommended number of readings was taken, for six different hypertension guidelines.
Objective To determine the effect of different hypertension management guidelines and of basing diagnosis on a single reading of blood pressure on the hypertension prevalence in the Indian population. Methods We performed a secondary analysis of data acquired as part of the Fourth national family health survey, 2015 to 2016, over all districts in India. We calculated the proportion of the population within three different age groups (18 to 34, 35 to 49 and 18 to 49 years of age) with raised blood pressure according to six different guidelines, and how prevalence changed if diagnoses were based on a single blood pressure measurement. Findings We observed that the Government of India and the American College of Cardiology/American Heart Association guidelines consistently yielded the lowest and highest prevalence of raised blood pressure; in the combined age group, we calculated the proportion of the population categorized as having raised blood pressure as 7.5% (95% confidence interval (CI): 7.4 to 7.7) and 40.1% (95% CI: 39.7 to 40.7), respectively. When basing diagnosis on a single reading of blood pressure only, a total of 56 million individuals would be erroneously categorized as hypertensive following the Government of India guidelines. We also showed that prevalence of hypertension in India varies with guidelines adhered to; in the combined age group, the national hypertension prevalence was three times higher when following the American College of Cardiology/American Heart Association compared with the Government of India guidelines. Conclusion To optimize current clinical practice, health-care providers need to follow universally agreed, evidence-based methods of diagnosing hypertension.

Data source
We used data from the most recent large-scale health survey, the Fourth National Family Health Survey, 24 conducted over 2015 to 2016 in India. The Fourth National Family Health Survey was conducted over all 640 districts of India (according to Census of India 2011 listing), 25 and included men aged 15 to 54 years and women aged 15 to 49 years. For consistency in our study, we included participants aged 18 to 49 years. Households within each district were selected to participate in the survey by two-stage cluster random sampling, stratified by rural versus urban areas. Primary sampling units, selected using probability proportional to population size, were defined as villages in rural areas and census enumeration blocks in urban areas. After sitting calmly for 5 minutes, the blood pressure of participants was measured three times, with at least 5 minutes between each measurement, in the left upper arm using the Omron HEM-8712 monitor. All blood pressure measurements were recorded in millimetres of mercury (mm Hg).

Hypertension guidelines
The six different guidelines that we used in this study for the calculation of hypertension prevalence are published by the European Society of Cardiology, 7 the Government of India, 8 the American College of Cardiology/American Heart Association, 9 the National Institute for Health and Clinical Excellence/British and Irish Hypertension Society, 10 the Eighth Joint National Committee 11 and the International Society of Hypertension. 12 The latter two guidelines are identical in terms of diagnosis of raised blood pressure. All six guidelines are used in India for the diagnosis and treatment of hypertension; although the exact proportion of health-care providers across India that adhere to any particular guideline is not known, the proportion of health-care providers adhering to specific guidelines was investigated within a single private hospital 26 and among attendees of a cardiology conference. 27 The number of readings taken, the number of visits required by the patient and the blood pressure at which hyper-tension is diagnosed (and pharmacological antihypertensive treatment initiated) for each of these six guidelines are listed in Table 1. Although all the listed guidelines recommend the measurement of blood pressure from at least two or three readings, it is general practice in India to take only a single blood pressure reading. 28 A patient is defined as having raised blood pressure if their blood pressure is categorized as Stage 1 or Grade 1 hypertension or higher according to the different guidelines in Table 1. A patient is defined as hypertensive if they have raised blood pressure, have confirmed at the time of the survey that they were taking prescribed medicine to control blood pressure, or if they had previously received at least two diagnoses of raised blood pressure or hypertension.

Statistical analysis
From the sample, we excluded participants having at least one missing blood pressure measurement or having unfeasible (i.e. systolic blood pressure < 30 mm Hg or < diastolic blood pressure) readings. We calculated the proportion of individuals within various blood pressure categories for age groups 18 to 34, 35 to 49 and 18 to 49 years of age for all six guidelines. We applied sampling weights and adjusted confidence intervals (CIs) at the primary sampling unit level to obtain nationally representative estimates with precise CIs.
To calculate prevalence, we defined participants having hypertension if they had stage I/grade I or higher blood pressure, taking prescribed medicine to control blood pressure or being informed at least twice by the health professional that they had raised blood pressure or hypertension. We estimated the total population within each age group from Census of India 2011 age distribution data, 25 multiplied by World Bank Indian population estimates for the year 2017. 29 We calculated the number of individuals across India within each category for each guideline by multiplying the proportion within each blood pressure category according to the Fourth National Family Health Survey by the calculated population within each age group. To confirm that the exclusion of participants did not cause significant difference in terms of age, sex and place of residence in the final data set, we per-formed a sensitivity analysis by comparing prevalence estimates from the clean data set with those from the full data set. We performed all analyses using Stata software, version 15.0 (StataCorp, College Station, United States of America).

Ethics
The Fourth National Family Health Survey obtained ethical clearance from the Ethics Committee of the International Institute for Population Sciences. 24 No specific permission was required for our study, as we conducted a secondary analysis of publicly available data.

Results
We obtained data on 797 161 individuals from the survey. We excluded 45 691 patients with missing data and 1594 participants with unusual blood pressure measurements. 24  We observe that the Government of India and the American College of Cardiology/American Heart Association guidelines consistently yield the lowest and highest prevalence of measured raised blood pressure, respectively ( the American College of Cardiology/ American Heart Association guidelines yielded proportions of the population with raised blood pressure of 3.4% (95% CI: 3.3 to 3.5) and 30.3% (95% CI: 30.2 to 30.5), respectively. Among the older age group, the lowest and highest proportions were calculated as 13.8% (95% CI: 13.7 to 14.0; Government of India guidelines) and 55.3% (95% CI: 55.2 to 55.5; American College of Cardiology/American Heart Association guidelines), respectively. Following the guidelines set by the Government of India, we estimate 48 million Indians have raised blood pressure; if the American College of Cardiology/American Heart Association guidelines are followed, this number is 253 million (Table 3).
We also observe an increase in the weighted proportion of the population classified as having raised blood pressure when only a single blood pressure reading (i.e. the first reading taken) is  considered, compared with measuring blood pressure from several readings as recommended by the guidelines (Table 2). If the proportion is based on first reading only, the guidelines published by the European Society of Cardiology/European Society for Hypertension and by the Government of India yield the same results. The increase in the proportion is higher in the younger compared with the older age group for all guidelines. Specifically, when we consider only the first reading for blood pressure categorization, the proportion of the population in the combined age group with raised blood pressure according to the Government of India guidelines increases by 8.9 percentage-points to 16.5% (95% CI: 16.5 to 16.7). According to the American College of Cardiology/American Heart Association guidelines, the proportion increases by 8.5 percentage-points to 48.6% (95% CI: 48.5 to 48.8) when only the first blood pressure reading is considered.
By neglecting to follow any guidelines precisely and by basing diagnosis on a single reading of blood pressure only, a total of 56 million would be erroneously categorized as hypertensive instead of normotensive following the American College of Cardiology/ American Heart Association guidelines ( Table 3). The largest increase in patients misdiagnosed with raised blood pressure from a single reading (65 million;

Discussion
This study compares the difference in hypertension prevalence when using six hypertension management guidelines in India. Our findings, that prevalence of hypertension varies according to guidelines followed and according to the number of blood pressure readings taken, are in concordance with other studies. 6 30 Our study is more versatile, however, with a comparison of six guidelines for hypertension prevalence using both single and repetitive measurements.
Our study had several limitations. We may have overestimated hypertension prevalence by our definition of hypertension being based on blood pressure measurements taken during one occasion; a clinical diagnosis of hypertension requires raised blood pressure on at least two different occasions. 32 The lower age of participants in this sample is also largely responsible for the lower hypertension prevalence observed here compared with the nationally representative study among an older sample. 18 Another limitation of our study is the lack of nationally representative data regarding use of hypertension guidelines by health-care providers in India. Finally, the questions asked in the Fourth National Family Health Survey did not allow us to investigate any connection between the prevalence of hypertension and lifestyle.
Our results show that the current use of several different guidelines in India results in inconsistent prevalence data, which could result in poor health outcomes. We therefore urge global bodies to discuss and propose a universal Blood pressure category by guideline  Framingham risk score (risk of developing cardio vascular disease within 10 years) of ≥ 10% as well as systolic blood pressure ≥ 130 mm Hg or diastolic blood pressure ≥ 80 mm Hg.
(. . .continued) guideline, similar to the cut-off for body mass index, malnutrition and anaemia. In our opinion, the European Society of Hypertension guidelines are most suited for India; these guidelines have the same definitions of blood pressure categories as the Government of India guidelines, but diagnosis is made from the last two readings (out of three) instead of the lowest reading (out of two or three). This recommendation is supported by two different studies. 30,33 To optimize current clinical practice in India, health-care providers need to follow universally agreed, evidencebased methods of diagnosing hypertension. The importance of multiple measurements and its impact on health management must be emphasized to health-care professionals. Once such guidelines have been agreed upon, their display at prominent locations within hospitals could help to improve the health literacy of the general population. ■ Acknowledgements MD and SR contributed equally to this work.
Funding: Ashish Awasthi is supported by the Department of Science and Technology, Government of India, New Delhi through INSPIRE Faculty program. No financial assistance was received in support of this study.
Competing interests: None declared.