Simplified hypertension screening methods across 60 countries: An observational study.

BACKGROUND: Simplified blood pressure (BP) screening approaches have been proposed. However, evidence is limited to a few countries and has not documented the cardiovascular risk amongst missed hypertension cases, limiting the uptake of these simplified approaches. We quantified the proportion of missed, over-diagnosed, and consistently identified hypertension cases and the 10-year cardiovascular risk in these groups. METHODS AND
FINDINGS: We used 60 WHO STEPS surveys (cross-sectional and nationally representative; n = 145,174) conducted in 60 countries in 6 world regions between 2004 and 2019. Nine simplified approaches were compared against the standard (average of the last 2 of 3 BP measurements). The 10-year cardiovascular risk was computed with the 2019 World Health Organization Cardiovascular Risk Charts. We used t tests to compare the cardiovascular risk between the missed and over-diagnosed cases and the consistent hypertension cases. We used Poisson multilevel regressions to identify risk factors for missed cases (adjusted for age, sex, body mass index, and 10-year cardiovascular risk). Across all countries, compared to the standard approach, the simplified approach that missed the fewest cases was using the second BP reading if the first BP reading was 130-145/80-95 mm Hg (5.62%); using only the second BP reading missed 5.82%. The simplified approach with the smallest over-diagnosis proportion was using the second BP reading if the first BP measurement was ≥140/90 mm Hg (3.03%). In many countries, cardiovascular risk was not significantly different between the missed and consistent hypertension groups, yet the mean was slightly lower amongst missed cases. Cardiovascular risk was positively associated with missed hypertension depending on the simplified approach. The main limitation of the work is the cross-sectional design.
CONCLUSIONS: Simplified BP screening approaches seem to have low misdiagnosis rates, and cardiovascular risk could be lower amongst missed cases than amongst consistent hypertension cases. Simplified BP screening approaches could be included in large screening programmes and busy clinics.

Introduction

High blood pressure (BP) [1,2] is highly prevalent and a major risk factor for cardiovascular morbidity and mortality worldwide; it disproportionally affects low- and middle-income countries (LMICs) [3], where screening for hypertension remains suboptimal [4]. Even though there are effective interventions and treatments for hypertension [5,6], patients first need to be diagnosed, which in the most comprehensive scenario requires multiple BP measurements on separate occasions [7-9], and in the most pragmatic approach requires 3 BP measurements, taking the average of the last 2 (standard approach) [10,11]. Three BP measurements may still be challenging in resource-constrained settings because of workforce shortage, low follow-up of patients, poor literacy, and affordability. Therefore, fewer BP measurements being needed to diagnose hypertension would help in screening large populations (e.g., May Measurement Month [12,13]) while saving time and resources in resource-constrained settings. Attempts have been made to find simplified BP screening approaches, such as only taking a second BP measurement if the first one was above a given threshold [14]. However, because these simplified approaches [14] have been tested in only 3 countries and BP may vary between countries [1], it is unknown whether there would be good agreement between the standard and the simplified approaches for BP screening in diverse populations. Furthermore, there may be concerns about the cardiovascular risk profile of hypertension cases missed by the simplified approaches, that is, whether the simplified approaches are missing people at high cardiovascular risk who would benefit from antihypertensive medication and cardiovascular disease prevention. To advance the literature on simplified BP screening approaches with data from multiple countries and to characterise the cardiovascular risk profile of the cases missed by the simplified approaches—with the aim of strengthening the recommendations for simplified BP screening approaches—we analysed national surveys in 60 countries. We aimed to answer the following research questions: What are the misdiagnosis and over-diagnosis rates for 9 simplified BP screening approaches? And what is the underlying cardiovascular risk profile for misdiagnosed, over-diagnosed, and consistently diagnosed cases?

Methods

Data sources and study population

We analysed WHO STEPS surveys [10,15]. These are population-based surveys conducted in nationally representative samples. These surveys follow a standard questionnaire and protocol including anthropometric, BP (S16 Table), and biomarker measurement [10]. If there were multiple surveys in a country, we used only the most recent one (i.e., only 1 survey per country was analysed).

We selected surveys with 3 BP measurements and with information on current smoking status, body mass index (BMI), waist circumference, fasting plasma glucose, and total cholesterol. We included adults aged 18–69 years and free of known hypertension (self-reported medical history and antihypertensive medication) [14]. In other words, we excluded people with hypertension because the simplified approaches for BP measurement are meant to be used in large screening programmes targeting the general population with undiagnosed hypertension.

None of the analyses presented in this paper were prespecified in a protocol. This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist).

Variable definition

The standard approach for BP screening uses the mean of the second and third BP measurements and defines hypertension as systolic BP ≥ 140 mm Hg or diastolic BP ≥ 90 mm Hg. This definition was compared to 9 simplified BP screening approaches: (i) first BP measurement; (ii) second BP measurement; (iii) average of the first and second BP measurements; (iv) second BP measurement if the first systolic BP is ≥130 mm Hg or the first diastolic BP is ≥80 mm Hg (otherwise the first BP measurement is used); (v) second BP measurement if the first systolic BP is ≥135 mm Hg or the first diastolic BP is ≥85 mm Hg (otherwise the first BP measurement is used); (vi) second BP measurement if the first systolic BP is ≥140 mm Hg or the first diastolic BP is ≥90 mm Hg (otherwise the first BP measurement is used); (vii) second BP measurement if the first systolic BP is 130–145 mm Hg or the first diastolic BP is 80–95 mm Hg (otherwise the first BP measurement is used); (viii) second BP measurement if the first systolic BP is 130–150 mm Hg or the first diastolic BP is 80–100 mm Hg (otherwise the first BP measurement is used); and (ix) second BP measurement if the first systolic BP is 130–155 mm Hg or the first diastolic BP is 80–105 mm Hg (otherwise the first BP measurement is used).

We compared each of the 9 simplified approaches with the standard to define 4 groups: (i) missed diagnosis: hypertension with the standard but non-hypertension with the simplified approach; (ii) over-diagnosis: non-hypertension with the standard but hypertension with the simplified approach; (iii) consistent hypertension: hypertension with both the standard and the simplified approach; and (iv) consistent non-hypertension: non-hypertension with both the standard and the simplified approach. In simple terms, misdiagnosis refers to people who would have been diagnosed with the standard approach but were not with the simplified approach; similarly, over-diagnosis refers to people who would not have been diagnosed with the standard approach but were with the simplified approach.

The proportion of missed hypertension cases was defined as the number of missed cases divided by the number of missed cases plus the number of consistent hypertension cases. The proportion of over-diagnosed cases was defined as the number of over-diagnosed cases divided by the number of over-diagnosed cases plus the number of consistent non-hypertension cases.

To characterise the cardiometabolic profile of the missed and over-diagnosed hypertension cases, we calculated the 10-year cardiovascular risk with the 2019 World Health Organization Cardiovascular Risk Charts [16]; we used the Stata package developed by the University of Cambridge Cardiovascular Epidemiology Unit [17]. We did not intend to make projections about cardiovascular risk; rather, we used 10-year cardiovascular risk as a summary measure to characterise overall cardiometabolic profile according to the groups of interest. Alternatively, we would have had to describe each cardiometabolic risk factor, making the results cumbersome. To compute 10-year predicted absolute cardiovascular risk, we used the same predictors as in the original 2019 WHO Cardiovascular Risk Charts except for diabetes; the original model included history of diabetes, whereas we included history and new diabetes cases (i.e., aware and unaware cases, the latter defined with fasting plasma glucose ≥ 126 mg/dl or 7 mmol/L). We included new diabetes cases in characterising cardiovascular risk profile in the missed and over-diagnosed populations because, in LMICs, many people with diabetes are unaware of their diagnosis [18]. We considered 10-year predicted cardiovascular risk as a continuous variable (from 0 to 1; not categorised as low/high risk).

Statistical analysis

This is a descriptive analysis conducted with R (version 4.1.1). Analysis scripts are available as S1 Cleaning and S1 Analysis and S1 WHO CVD Risk. A p-value ≤ 0.05 was considered statistically significant. We did not use sampling weights in the analyses because we aimed to compare groups of simplified BP screening approaches rather than reporting prevalences representative at the country level [14].

First, we summarised the proportions (%) of missed and over-diagnosed cases at the global, regional, and country level (countries in each region are listed in S1 Table). Second, we summarised mean 10-year cardiovascular risk according to the 4 groups: missed diagnosis, over-diagnosis, consistent hypertension, and consistent non-hypertension. We used t tests to compare the mean cardiovascular risk in the missed, over-diagnosed, and consistent non-hypertension groups against that in the consistent hypertension group. The p-values for these t tests are reported for crude analyses and for analyses adjusted for multiple comparisons (Bonferroni). Third, to identify potential correlates of being misdiagnosed, we fitted individual-level Poisson multilevel regression models. The outcome was misdiagnosis according to each of the 9 simplified approaches (yes/no); thus, we had 9 regression models. The independent variables (i.e., predictors) were sex (reference was men), age (years), BMI (kg/m2), and 10-year cardiovascular risk. Random intercepts were included whereby countries were nested within regions. The regression results are presented as prevalence ratios (PRs) along with 95% confidence intervals (95% CIs).

Ethics

We analysed de-identified open-access data [15]. We did not request approval from an ethics committee. The authors alone are responsible for the opinions in the paper. The funder had no role in the study design, analysis, results interpretation, or conclusions. The first 2 authors conducted the analyses and vouch for the accuracy of the results.

Results

Data description

We analysed 60 STEPS surveys from 60 countries including 145,174 individuals (S1 Supplementary Flow Chart); the smallest sample size was 215 people in British Virgin Islands, and the largest was 7,431 people in Ethiopia (Table 1). The mean age ranged from 34 to 44 years, and the number of men ranged from 90 to 3,164. In all countries, the first BP record was higher than the average of the second and third records; the mean difference between the first and the average of the last 2 readings ranged from 0.70 mm Hg (Lebanon) to 5.75 mm Hg (Cambodia).

Table 1

Descriptive statistics of the analysed surveys.

Country	Year	Sample	Men	Age (years)	First SBP (mm Hg)	Second SBP (mm Hg)	Third SBP (mm Hg)	BMI (kg/m2)	Waist circumference (cm)	FPG (mmol/L)	TC (mmol/L)	10-year CV risk
Afghanistan	2018	2,550	1,503	36	127.1	124.0	122.6	24.7	86.3	5.1	3.8	4.1
Algeria	2017	4,653	2,234	39	129.0	125.8	124.7	26.4	92.3	5.4	4.2	5.0
American Samoa	2004	1,137	503	42	134.2	131.5	130.6	35.3	105.3	7.2	4.8	5.2
Armenia	2016	1,205	355	42	131.0	129.3	127.9	26.5	90.9	4.6	4.4	4.9
Azerbaijan	2017	1,772	782	42	126.5	124.5	122.9	26.6	89.9	5.1	4.4	4.3
Bangladesh	2018	5,688	2,773	38	121.0	118.5	117.3	22.7	78.7	5.4	4.4	2.6
Belarus	2017	2,744	1,253	41	130.5	128.6	127.5	25.8	85.9	4.6	4.7	5.8
Benin	2015	4,397	2,116	37	130.3	127.1	125.0	23.5	81.9	4.9	4.0	2.0
Bhutan	2019	4,149	1,622	39	125.5	123.2	122.1	25.2	82.7	4.6	3.7	2.2
Botswana	2014	2,231	811	34	127.9	125.1	123.9	23.8	80.6	4.5	3.7	1.8
British Virgin Islands	2009	215	90	41	129.7	127.2	126.2	28.6	91.0	5.9	4.8	NA
Brunei Darussalam	2016	1,242	526	37	124.1	120.4	119.3	26.5	85.1	5.1	5.1	2.8
Cabo Verde	2007	693	278	42	138.5	134.5	132.6	24.6	85.3	5.6	4.2	2.8
Cambodia	2010	4,403	1,647	42	120.1	115.0	113.7	21.7	75.4	4.0	4.5	2.3
Comoros	2011	1,139	280	41	133.2	128.5	126.8	26.3	88.6	4.1	4.6	2.5
Cook Islands	2015	567	245	39	131.5	127.8	125.6	33.5	103.1	6.3	5.0	NA
Ecuador	2018	3,172	1,361	39	120.7	118.1	117.0	27.0	89.1	5.2	4.4	2.0
Eritrea	2010	4,152	1,103	43	121.2	117.8	116.0	20.2	75.5	4.1	4.8	2.5
Eswatini	2014	1,759	701	36	127.5	124.8	122.8	26.0	84.3	5.1	3.8	2.0
Ethiopia	2015	7,431	3,164	35	123.1	120.5	119.2	20.8	75.5	4.5	3.6	1.5
Georgia	2016	1,855	551	44	128.2	125.5	124.0	27.9	90.8	4.5	4.4	5.0
Guyana	2016	572	227	38	125.2	123.4	122.4	26.4	91.8	5.0	5.1	2.8
Iraq	2015	3,166	1,301	39	131.5	130.0	129.7	29.3	97.7	5.9	4.8	6.5
Jordan	2019	2,320	843	38	118.3	114.7	114.3	28.2	90.7	4.6	3.9	4.5
Kenya	2015	3,463	1,462	37	128.0	124.8	123.0	23.1	78.7	4.7	3.7	1.9
Kiribati	2016	966	429	38	130.1	126.0	125.0	29.3	90.8	6.0	4.0	4.0
Kuwait	2014	1,461	576	35	118.9	117.5	117.2	29.1	89.3	5.5	5.0	4.0
Kyrgyzstan	2013	1,815	721	41	132.8	130.0	128.2	26.6	87.9	4.8	4.4	3.7
Lao People’s Democratic Republic	2013	2,156	880	38	119.6	116.5	115.6	22.7	77.2	4.5	4.2	2.1
Lebanon	2017	793	298	44	126.1	125.5	125.3	27.6	94.1	5.4	5.5	8.6
Lesotho	2012	1,493	552	41	130.6	127.2	125.6	25.5	83.1	4.5	3.5	2.5
Libya	2009	1,408	805	40	138.8	135.2	134.0	27.6	94.4	5.4	4.6	6.6
Malawi	2017	3,199	1,231	37	123.8	120.7	119.3	22.9	77.3	4.7	3.7	1.9
Mongolia	2019	4,019	1,899	39	121.8	119.3	118.1	25.7	86.7	5.8	4.4	3.3
Morocco	2017	3,540	1,282	41	130.8	127.8	126.3	26.4	93.1	5.6	3.6	5.4
Myanmar	2014	5,541	2,069	43	125.1	122.6	121.8	22.1	76.7	5.1	4.6	3.0
Nauru	2016	706	333	35	123.4	120.2	120.2	34.2	102.4	5.4	3.9	NA
Nepal	2019	4,428	1,546	40	127.1	125.0	123.8	23.0	80.5	5.2	3.7	3.0
Niue	2012	492	233	40	130.6	126.1	125.1	32.4	97.7	6.7	4.6	NA
Qatar	2012	1,057	422	36	118.2	116.5	115.4	29.3	94.0	5.1	4.2	3.7
Republic of Moldova	2013	2,095	784	42	133.2	130.3	128.5	26.4	86.0	5.1	4.5	6.1
Rwanda	2013	5,646	2,164	36	126.4	122.9	120.9	22.5	76.6	3.9	3.2	1.7
Samoa	2013	1,314	518	38	128.6	124.3	123.5	32.4	99.9	6.7	4.4	3.7
Sao Tome and Principe	2019	1,351	594	35	127.0	123.1	121.8	24.3	83.7	5.5	5.3	2.1
Seychelles	2004	849	396	42	128.0	124.9	122.7	26.4	88.2	5.7	5.4	3.2
Solomon Islands	2015	1,422	666	39	126.7	122.4	121.3	26.8	85.8	4.5	4.5	3.5
Sri Lanka	2015	3,181	1,294	42	127.3	125.1	123.9	23.0	82.4	4.7	4.1	3.1
Sudan	2016	5,692	2,110	37	131.9	127.8	126.3	23.3	83.6	4.7	3.9	4.2
Tajikistan	2017	1,924	855	38	134.3	130.4	128.2	25.5	82.0	5.2	3.9	2.8
Timor-Leste	2014	2,064	868	40	126.7	124.3	123.2	21.0	76.0	4.3	3.8	2.9
Togo	2011	1,186	546	36	128.0	124.3	123.3	23.2	80.7	4.3	4.4	1.7
Tokelau	2014	413	199	36	128.4	123.7	122.9	33.3	100.3	6.9	5.1	NA
Turkmenistan	2018	3,108	1,386	38	129.0	126.3	124.5	25.4	89.2	5.1	4.2	2.4
Tuvalu	2015	823	392	40	138.6	134.1	133.2	32.6	99.9	4.9	4.1	NA
Uganda	2014	2,976	1,287	35	129.3	124.8	122.7	22.7	78.5	4.0	3.5	1.6
United Republic of Tanzania	2012	1,555	717	42	136.4	132.8	130.6	22.8	83.1	4.8	4.5	3.1
Uruguay	2014	875	313	41	125.2	123.9	123.0	26.6	89.9	5.1	4.6	3.0
Vanuatu	2011	3,643	1,865	41	134.5	130.6	129.5	26.1	77.0	5.7	4.9	4.1
Vietnam	2015	2,517	1,076	42	122.6	119.3	118.0	22.0	76.6	4.0	4.5	2.8
Zambia	2017	2,791	1,135	35	127.0	124.2	122.3	23.0	79.1	5.0	3.4	1.9

Sample and men are absolute numbers. All other numeric variables are presented as means. Ten-year CV risk is 10-year predicted CV risk based on the 2019 World Health Organization Cardiovascular Risk Charts [16]. There are 6 countries with missing information for 10-year predicted CV risk because the risk model did not provide results for these countries (British Virgin Islands, Cook Islands, Niue, Nauru, Tokelau, and Tuvalu). The standard deviations for the numeric variables presented in this table are shown in S15 Table. BMI, body mass index; CV, cardiovascular; FPG, fasting plasma glucose; NA, not available; SBP, systolic blood pressure; TC, total cholesterol.

Misdiagnosis

Across all countries, compared to the standard approach, the simplified approach that missed the fewest cases was using the second BP reading if the first BP measurement was 130–145/80–95 mm Hg; for this simplified approach, the misdiagnosed proportion was 5.62% (Table 2). The approach of using only the second BP reading missed 5.82% of cases. The other simplified approaches missed more than 6% of hypertension cases; the simplified approach that missed the most cases was using the second BP reading if the first BP measurement was ≥140/90 mm Hg (15.15% misdiagnosis).

Table 2

Proportion (%) of missed hypertension cases for each of the 9 simplified approaches: Global results.

Simplified approach	Mean (%)	SD	Median (%)	Min (%)	Max (%)
1st BP record	10.18	3.45	10.08	2.21	18.88
2nd BP record	5.82	2.41	5.56	1.21	12.50
Average of 1st and 2nd BP records	7.26	2.31	6.88	2.65	13.24
2nd BP record if 1st BP record ≥ 130/80	7.38	3.01	6.91	1.21	15.34
2nd BP record if 1st BP record ≥ 135/85	9.57	3.40	9.04	2.92	17.70
2nd BP record if 1st BP record ≥ 140/90	15.15	4.88	14.71	3.45	28.02
2nd BP record if 1st BP record = 130–145/80–95	5.62	2.44	5.35	1.21	13.86
2nd BP record if 1st BP record = 130–150/80–100	6.65	2.75	6.33	1.21	15.34
2nd BP record if 1st BP record = 130–155/80–105	7.07	2.91	6.73	1.21	15.34

All estimates shown in the tables (mean, SD, median, minimum, and maximum) are across the 60 countries included in the analysis. BP values given in millimetres of mercury. BP, blood pressure; max, maximum; min, minimum; SD, standard deviation.

The same pattern was observed at the region level (Table 3). In 4 out of the 6 regions, the smallest proportion of misdiagnosis was found with the second BP record if the first BP measurement was 130–145/80–95 mm Hg: from 3.64% (Europe) to 9.25% (Americas). In the 2 remaining regions, the smallest proportion of misdiagnosis was found with the second BP reading only: 4.99% in the Eastern Mediterranean and 5.40% in Africa. Of note, across all regions the largest misdiagnosis proportion was found when using the second BP reading if the first BP measurement was ≥140/90 mm Hg.

Table 3

Proportion (%) of missed hypertension cases for each of the 9 simplified approaches by WHO world region.

Region	Simplified approach	Mean (%)	SD	Median (%)	Min (%)	Max (%)
Africa	1st BP record	10.25	3.22	10.23	5.56	14.48
	2nd BP record	5.45	1.84	5.48	2.88	8.68
	Average of 1st and 2nd BP records	7.34	2.14	7.21	3.85	10.34
	2nd BP record if 1st BP record ≥ 130/80	7.18	2.48	6.62	3.63	11.32
	2nd BP record if 1st BP record ≥ 135/85	9.51	3.19	9.75	4.70	14.66
	2nd BP record if 1st BP record ≥ 140/90	15.07	4.13	15.29	8.76	20.91
	2nd BP record if 1st BP record = 130–145/80–95	5.52	1.92	5.37	2.56	9.06
	2nd BP record if 1st BP record = 130–150/80–100	6.50	2.19	6.04	3.21	10.19
	2nd BP record if 1st BP record = 130–155/80–105	6.82	2.24	6.52	3.42	10.19
Americas	1st BP record	14.90	2.82	14.11	12.50	18.88
	2nd BP record	9.85	2.52	9.71	7.49	12.50
	Average of 1st and 2nd BP records	10.38	1.93	9.81	8.91	12.98
	2nd BP record if 1st BP record ≥ 130/80	11.77	3.54	11.60	8.56	15.34
	2nd BP record if 1st BP record ≥ 135/85	14.62	2.73	14.45	11.88	17.70
	2nd BP record if 1st BP record ≥ 140/90	23.32	3.61	22.99	19.25	28.02
	2nd BP record if 1st BP record = 130–145/80–95	9.25	3.87	9.10	4.95	13.86
	2nd BP record if 1st BP record = 130–150/80–100	11.39	3.97	11.15	7.92	15.34
	2nd BP record if 1st BP record = 130–155/80–105	11.64	3.71	11.60	8.02	15.34
Eastern Mediterranean	1st BP record	8.91	4.12	9.31	2.21	15.22
	2nd BP record	4.99	2.53	5.53	1.21	8.26
	Average of 1st and 2nd BP records	6.64	2.42	6.88	2.65	9.97
	2nd BP record if 1st BP record ≥ 130/80	6.49	3.22	7.10	1.21	10.88
	2nd BP record if 1st BP record ≥ 135/85	8.72	3.61	8.32	2.92	14.42
	2nd BP record if 1st BP record ≥ 140/90	13.37	6.63	13.99	3.45	23.91
	2nd BP record if 1st BP record = 130–145/80–95	5.12	2.22	5.51	1.21	7.25
	2nd BP record if 1st BP record = 130–150/80–100	5.94	2.84	6.26	1.21	9.42
	2nd BP record if 1st BP record = 130–155/80–105	6.30	3.04	6.75	1.21	9.96
Europe	1st BP record	8.91	1.86	9.08	6.07	11.79
	2nd BP record	3.83	0.60	3.96	3.03	4.62
	Average of 1st and 2nd BP records	5.59	1.30	5.90	3.56	7.51
	2nd BP record if 1st BP record ≥ 130/80	4.79	0.70	4.66	3.85	5.86
	2nd BP record if 1st BP record ≥ 135/85	6.59	0.91	6.52	5.36	7.81
	2nd BP record if 1st BP record ≥ 140/90	11.90	1.68	11.55	10.04	14.11
	2nd BP record if 1st BP record = 130–145/80–95	3.64	0.68	3.85	2.51	4.50
	2nd BP record if 1st BP record = 130–150/80–100	4.36	0.85	4.66	3.14	5.26
	2nd BP record if 1st BP record = 130–155/80–105	4.47	0.85	4.66	3.14	5.41
Southeast Asia	1st BP record	8.89	2.55	8.38	5.07	12.12
	2nd BP record	4.26	0.76	4.07	3.56	5.56
	Average of 1st and 2nd BP records	6.34	1.55	6.22	4.21	8.63
	2nd BP record if 1st BP record ≥ 130/80	5.11	1.09	4.64	4.12	6.81
	2nd BP record if 1st BP record ≥ 135/85	6.64	1.39	6.63	4.81	8.40
	2nd BP record if 1st BP record ≥ 140/90	12.14	3.06	11.03	8.42	16.31
	2nd BP record if 1st BP record = 130–145/80–95	4.25	1.10	4.04	3.01	5.90
	2nd BP record if 1st BP record = 130–150/80–100	4.83	1.16	4.43	3.61	6.70
	2nd BP record if 1st BP record = 130–155/80–105	5.01	1.10	4.50	4.12	6.81
Western Pacific	1st BP record	10.80	3.43	10.66	5.67	17.65
	2nd BP record	7.29	1.96	7.38	3.62	10.76
	Average of 1st and 2nd BP records	7.89	2.38	7.12	4.18	13.24
	2nd BP record if 1st BP record ≥ 130/80	9.17	2.35	9.20	4.74	13.24
	2nd BP record if 1st BP record ≥ 135/85	11.41	2.41	11.67	7.80	16.91
	2nd BP record if 1st BP record ≥ 140/90	16.96	3.17	17.32	10.86	22.79
	2nd BP record if 1st BP record = 130–145/80–95	6.59	2.40	6.37	2.70	11.03
	2nd BP record if 1st BP record = 130–150/80–100	7.83	2.10	7.63	4.46	11.03
	2nd BP record if 1st BP record = 130–155/80–105	8.69	2.35	8.66	4.46	13.24

BP values given in millimetres of mercury. BP, blood pressure; max, maximum; min, minimum; SD, standard deviation; WHO, World Health Organization.

At the country level, the same pattern arose (Fig 1). In general, the smallest proportions of misdiagnosis were found with the second BP reading if the first BP measurement was 130–145/80–95 mm Hg or with the second BP measurement only. For both simplified approaches, Kuwait had the smallest proportion of misdiagnosis (1%), whilst Ecuador had the largest (14% and 12%, respectively). Notably, the misdiagnosis proportions were high when using the second BP reading if the first BP measurement was ≥140/90 mm Hg.

Fig 1

Proportion (%) of missed hypertension cases for the 9 simplified approaches, stratified by country and region.

BP values given in millimetres of mercury. BP, blood pressure.

Overall, based solely on misdiagnosis, using the second BP record if the first BP measurement is 130–145/80–95 mm Hg, or the second BP reading only, seem to be reasonable simplified approaches. Conversely, using the second BP reading if the first BP measurement is ≥140/90 mm Hg does not appear to be a reasonable option because it yields large misdiagnosis rates.

Over-diagnosis

Globally (S2 Table), as well as across the 6 regions (S3 Table), the simplified approach with the smallest over-diagnosis proportion was using the second BP reading if the first BP measurement was ≥140/90 mm Hg. Globally, this proportion was 3.03%; across regions this proportion varied between 1.92% (Americas) and 3.77% (Europe). The over-diagnosis proportions by region for the simplified approach based on the second BP reading only were around 5%: 4.48% (Americas), 5.05% (Southeast Asia), 5.65% (Western Pacific), 5.71% (Eastern Mediterranean), 6.11% (Africa), and 6.86% (Europe). The over-diagnosis proportions by region for the simplified approach based on the second BP reading if the first BP measurement was 130–145/80–95 mm Hg were as low as 6.71% (Southeast Asia) and 6.79% (Americas), and as high as 9.61% (Europe) and 9.68% (Africa).

At the country level, the same pattern emerged (S1 Fig). The over-diagnosis proportion based on the second BP reading if the first BP measurement was ≥140/90 mm Hg was 1% in 7 countries. The over-diagnosis proportion based on the second BP reading alone was lowest in Kuwait and Cambodia (1%), whereas it was largest in Kyrgyzstan and Tajikistan (11%). The over-diagnosis proportion based on the second BP if the first BP measurement was 130–145/80–95 mm Hg was lowest in Kuwait (3%) and highest in Tajikistan (17%).

Even though the simplified approach of using the second BP record if the first BP measurement was ≥140/90 mm Hg yielded the lowest proportion of over-diagnosis, this approach had the largest proportion of misdiagnosis (as detailed in the previous section). The simplified approach based on the second BP reading alone had reasonable proportions of over-diagnosis and low rates of misdiagnosis (as detailed in the previous section).

Cardiovascular risk profile amongst the misdiagnosed cases

Six countries (n = 3,216) did not have data for 10-year cardiovascular risk; thus, there were 54 countries (n = 141,958) included in this analysis. Descriptive statistics of the assessed risk factors per survey and stratified by the simplified approaches are available in S4 Table. The distribution of 10-year cardiovascular risk by country (hence survey), shown in S1 Fig, suggests there were no outliers.

Regarding the simplified approach based on the second BP measurement only, there were 34 countries where mean 10-year cardiovascular risk was not different between the missed and consistent hypertension groups (S3 Fig), yet the mean 10-year cardiovascular risk was slightly lower in the former (mean = 4.17, SD = 1.78) than the latter (mean = 6.32, SD = 2.56) group. Moreover, higher 10-year cardiovascular risk was not associated with higher prevalence of being misdiagnosed (Table 4). The mean 10-year cardiovascular risk in the missed hypertension group was 21.64% of the mean in the consistent hypertension group in Azerbaijan, and 98.45% in Iraq (S14 Table). This suggests that in countries where there was no difference in the mean 10-year cardiovascular risk (e.g., Iraq; S3 Fig), the mean 10-year cardiovascular risk was very similar between the missed and consistent hypertension groups.

Table 4

Multilevel regression models for missed hypertension diagnosis according to the 9 simplified BP screening approaches.

Simplified approach and independent variable	PR (95% CI)	p-Value
1st BP record (n = 141,958)
Female sex	0.92 (0.85–0.99)	0.028
Age	1.01 (1.00–1.01)	0.001
BMI	1.02 (1.02–1.03)	<0.001
Cardiovascular risk	2.78 (1.03–7.46)	0.043
Random effects, country: region (variance)	0.09
Random effects, region (variance)	0.00
2nd BP record (n = 141,958)
Female sex	0.88 (0.79–0.97)	0.011
Age	1.01 (1.00–1.01)	0.012
BMI	1.03 (1.02–1.04)	<0.001
Cardiovascular risk	2.51 (0.67–9.43)	0.172
Random effects, country: region (variance)	0.10
Random effects, region (variance)	0.00
Average of 1st and 2nd BP records (n = 141,958)
Female sex	0.88 (0.80–0.96)	0.003
Age	1.01 (1.00–1.01)	0.001
BMI	1.03 (1.02–1.04)	<0.001
Cardiovascular risk	0.88 (0.25–3.09)	0.841
Random effects, country: region (variance)	0.08
Random effects, region (variance)	0.01
2nd BP record if 1st BP record ≥ 130/80 (n = 141,958)
Female sex	0.91 (0.83–0.99)	0.034
Age	1.00 (1.00–1.01)	0.061
BMI	1.03 (1.02–1.04)	<0.001
Cardiovascular risk	3.20 (0.98–10.41)	0.054
Random effects, country: region (variance)	0.11
Random effects, region (variance)	0.00
2nd BP record if 1st BP record ≥ 135/85 (n = 141,958)
Female sex	0.89 (0.82–0.97)	0.006
Age	1.01 (1.00–1.01)	0.019
BMI	1.03 (1.02–1.03)	<0.001
Cardiovascular risk	3.57 (1.29–9.88)	0.0145
Random effects, country: region (variance)	0.09
Random effects, region (variance)	0.00
2nd BP record if 1st BP record ≥ 140/90 (n = 141,958)
Female sex	0.89 (0.84–0.95)	<0.001
Age	1.01 (1.01–1.01)	<0.001
BMI	1.03 (1.02–1.04)	<0.001
Cardiovascular risk	2.59 (1.14–5.84)	0.022
Random effects, country: region (variance)	0.10
Random effects, region (variance)	0.00
2nd BP record if 1st BP record = 130–145/80–95 (n = 141,958)
Female sex	0.92 (0.83–1.02)	0.123
Age	1.00 (1.00–1.01)	0.208
BMI	1.03 (1.02–1.03)	<0.001
Cardiovascular risk	2.42 (0.60–9.72)	0.212
Random effects, country: region (variance)	0.09
Random effects, region (variance)	0.00
2nd BP record if 1st BP record = 130–150/80–100 (n = 141,958)
Female sex	0.89 (0.81–0.98)	0.015
Age	1.00 (1.00–1.01)	0.296
BMI	1.03 (1.02–1.04)	<0.001
Cardiovascular risk	2.69 (0.74–9.76)	0.132
Random effects, country: region (variance)	0.09
Random effects, region (variance)	0.00
2nd BP record if 1st BP record = 130–155/80–105 (n = 141,958)
Female sex	0.90 (0.82–0.98)	0.021
Age	1.00 (1.00–1.01)	0.160
BMI	1.03 (1.02–1.04)	<0.001
Cardiovascular risk	2.59 (0.74–9.01)	0.136
Random effects, country: region (variance)	0.09
Random effects, region (variance)	0.00

In all regression models the outcome was misdiagnosed (yes/no) for each simplified approach in comparison to the standard (average of the second and third BP measurements). The regression models included all independent variables together (i.e., only adjusted models were computed). BP values given in millimetres of mercury. 95% CI, 95% confidence interval; BP, blood pressure; BMI, body mass index; PR, prevalence ratio.

Regarding the simplified approach of using the second BP reading if the first BP measurement is 130–145/80–95 mm Hg, there were 34 countries where mean 10-year cardiovascular risk was not different between the missed and consistent hypertension groups (S8 Fig), yet the mean was slightly lower in the former (mean = 4.00; SD = 1.06) than the latter (mean = 6.13; SD = 2.40) group. Higher 10-year cardiovascular risk was not associated with higher prevalence of being misdiagnosed (Table 4). The mean 10-year cardiovascular risk in the missed hypertension group was 20.51% of the mean in the consistent hypertension group in Azerbaijan, and 94.17% in Iraq (S14 Table). This suggests that in countries where there was no difference in the mean 10-year cardiovascular risk (e.g., Iraq; S8 Fig), the mean cardiovascular risk was very similar between the consistent and missed hypertension groups (i.e., a ratio close to 94%).

These comparisons (including sampling sizes and p-values) for all of the simplified approaches are available in S2–S10 Figs and S5–S13 Tables.

The fact that mean cardiovascular risk was not significantly different between the missed and consistent hypertension groups for the 2 best-performing simplified approaches agrees with the results of the regression models, in which 10-year predicted risk was not associated with higher prevalence of misdiagnosis in these 2 simplified approaches. Together, these results may imply that in-country evaluations of these simplified approaches are needed to determine whether the missed cases have higher cardiovascular risk or not.

Cardiovascular risk profile amongst the over-diagnosed cases

Regarding the simplified approach based on the second BP measurement only, there were 10 countries where mean 10-year cardiovascular risk was not different between the missed and consistent hypertension groups (S3 Fig), yet the mean 10-year cardiovascular risk was slightly lower in the former (mean = 5.43, SD = 2.20) than the latter (mean = 7.33, SD = 2.79) group. The mean 10-year cardiovascular risk in the missed group was 30.89% of the mean in the consistent hypertension group in Armenia, and 104.85% in Kuwait (S14 Table). This suggests that in countries where there was no difference in the mean 10-year cardiovascular risk (e.g., Kuwait; S3 Fig), the mean cardiovascular risk was similar between the missed and consistent hypertension groups.

Regarding the simplified approach of using the second BP reading if the first BP measurement is 130–145/80–95 mm Hg, there were 6 countries where mean 10-year cardiovascular risk was not different between the missed and consistent hypertension groups (S8 Fig), yet the mean was slightly smaller in the former (mean = 6.33; SD = 2.18) than the latter (mean = 8.17; SD = 3.25) group. The mean 10-year cardiovascular risk in the missed hypertension group was 35.41% of the mean in the consistent hypertension group in Zambia, and 99.87% in Kuwait (S14 Table). This suggests that in countries where there was no difference in the mean 10-year cardiovascular risk (e.g., Kuwait; S8 Fig), the mean cardiovascular risk was very similar between the consistent and missed hypertension groups (i.e., a ratio close to 100%).

Potential correlates for missed hypertension cases

Age and BMI were positively associated with a higher probability of being misdiagnosed in all 9 simplified BP screening approaches, though with a small magnitude: The PR ranged between 1.00 and 1.03 (Table 4). Conversely, female sex (in comparison to male sex) was associated with a lower probability of being misdiagnosed with most of the 9 simplified approaches. Even though the associations for these independent variables were statistically significant, the strength of the associations may be negligible.

Ten-year cardiovascular risk showed a mixed profile, yielding strong positive associations for some of the simplified approaches. Overall, the simplified approaches in which 10-year cardiovascular risk was associated with higher prevalence of misdiagnosis could be less optimal. Conversely, simplified approaches in which 10-year cardiovascular risk did not show a strong association could warrant further attention, and be further considered.

The variability of the regression models was always larger between countries within regions, than between regions (consistent with Fig 1). This may imply that country-specific guidelines are needed for following simplified approaches, rather than 1 guideline for all countries in a region.

Discussion

Main findings

Leveraging 60 national surveys, we documented concordance between hypertension diagnosis based on the average of the last 2 of 3 BP measurements (standard approach) and 9 simplified approaches (e.g., second BP measurement if the first was above a threshold). The proportion of missed cases was lowest when using the second BP reading if the first BP measurement was 130–145/80–95 mm Hg, followed by using the second BP only. Notably, the former simplified approach would require a second BP measurement in some people only, reducing the total number of BP measurements, and therefore time and resources used, which could allow more people to be screened. We observed differences between countries within world regions. Also, we quantified the absolute cardiovascular risk in the missed hypertension group. In many countries, the mean cardiovascular risk was not different between the missed hypertension and consistent hypertension groups, yet the mean cardiovascular risk in the missed group was slightly lower than that in the consistent hypertension group.

Altogether, this research shows that simplified BP screening approaches may be sensible and could increase the number of people screened for hypertension, particularly in LMICs where screening for hypertension is still limited [4]. However, it would seem reasonable not to have a one-size-fits-all simplified approach. Although physicians may have reasonable concerns about missing hypertension cases with the simplified approaches, our findings suggest that missed cases may have slightly lower absolute cardiovascular risk than their peers with hypertension. Also, cardiovascular risk was positively associated with missed hypertension for only some simplified approaches. Future work with prospective cohorts should confirm this observation before simplified approaches are strongly recommended.

Across all countries, the proportion of missed hypertension in our study was similar to the proportion reported for simplified screening approaches in the US [14]: 10.2% versus 9.6% [14] for the first BP record, 5.8% versus 4.9% [14] for the second BP record, 7.3% versus 7.2% [14] for the average of the first and second BP records, and 7.4% versus 5.2% [14] for the second BP record if the first was ≥130/80 mm Hg. Conversely, our proportions of over-diagnosis were more than 2 times the proportions in the US: 14.4% versus 4.3% [14] for the first BP record, 5.8% versus 2.0% for the second BP record, 7.4% versus 2.0% for the average of the first 2 BP records, and 5.1% versus 2.0% [14] for the second BP record when the first was ≥130/80 [14]. The similar proportions for missed hypertension may suggest that the simplified BP screening approaches are sensible and little biased by measurement protocols. The higher over-diagnosis found in our study could be owing to different BP measurement protocols between the STEPS surveys and the US national health survey [10,14]. Arguably, over-diagnosis would not be an unfavourable outcome, particularly when antihypertensive treatment is initiated at lower BP thresholds (provided the patient has other indications like history of cardiovascular disease or high cardiovascular risk) [7,9].

The results highlight that some of the 9 simplified approaches may lead to little misdiagnosis and are not associated with higher overall cardiovascular risk; however, we cautiously believe that further validation of these simplified approaches is warranted. Large prospective studies are needed to study the long-term cardiovascular outcomes for each simplified approach. Nevertheless, we would cautiously suggest considering 2 simplified approaches: (i) using only the second BP measurement and (ii) using the second BP reading if the first BP measurement is 130–145/80–95 mm Hg. If deemed necessary by local experts, these 2 simplified approaches could be implemented in screening programmes. Also, these 2 simplified approaches could be subject of further in-country validation analyses.

The WHO STEPS protocol [10], like other similar guidelines, recommends waiting 3 minutes between BP measurements. If there are 3 measurements (standard), and we assume that each measurement takes seconds, then measuring BP in 1 person could take at least 6 minutes. This would be equivalent to measuring BP in 10 people per hour. However, if a simplified approach is implemented, whereby, for example, only 2 measurements are required, the time invested to measure BP in 1 person would be approximately 3 minutes. In other words, we could measure BP in 20 people per hour, substantially increasing the number of individuals who could be screened for hypertension. The simplified approaches could save 50% of the time needed to measure BP in 1 person compared to current and standard guidelines.

Therefore, the potential applications of our work target several relevant scenarios. First, our work could influence May Measurement Month [12,13]. This is a global hypertension screening programme conducted yearly, and since 2016 it has covered more than 100 countries, benefiting over 1,000,000 people. This programme follows the 3-measurement protocol. Our work could inform future May Measurement Month campaigns by motivating discussion on whether fewer BP measurements could be taken, to maximise resources while reaching a much larger population. Second, our work could also influence future research and large health surveys. In addition to being used in the WHO STEPS surveys themselves, the WHO STEPS survey protocol has influenced other population-based health surveys worldwide, which would also take 3 BP measurements. Our work could spark interest in discussing whether 3 BP measurements are needed, or whether taking fewer measurements is a reasonable option to save resources that could be used to measure other relevant health variables. Third, in some clinics there may be a lack of sphygmomanometers or a shortage of personnel, limiting the number of people who can be screened for hypertension. Our work could deliver pragmatic approaches to optimise the protocols for BP measurement, to maximise the number of people who can be screened.

Our results showed large variability across countries within world regions. While simplified BP screening approaches may be a sensible and pragmatic alternative for screening large populations, a one-size-fits-all simplified approach may not be possible. Countries may need to find the optimal trade-off between the number of BP measurements and hypertension cases missed. Health organisations could set protocols for each country to define simplified BP screening approaches, so that these can be used in massive screening programmes [12,13].

Ten-year cardiovascular risk was positively associated with missed hypertension cases depending on the simplified approach used. Based on this, our results do not support relying on the first BP measurement only, for example. Conversely, our work may support using the second BP measurement or the average of the first 2 measurements, because 10-year cardiovascular risk was not associated with misdiagnosis amongst cases missed by these simplified approaches. The regression coefficients for absolute cardiovascular risk in some of the models, depending on the simplified approach, did not reach statistical significance. We would argue that this signalled groups of missed cases in which cardiovascular risk was not (truly) associated with misdiagnosis. Given the large sample size included in the models, the results most likely show strong associations (or lack of association) than unstable results. Nevertheless, residual confounding could still be a possibility, even though we included the relevant and potential confounders that were available. Our results must be further validated with larger samples and, more importantly, with prospective cohort studies to examine the mid- and long-term cardiovascular outcomes of the missed and consistent hypertension groups [19].

Strengths and limitations

Our study advances current knowledge on simplified BP screening approaches with estimates from 60 LMICs, and describes 10-year cardiovascular risk in the missed diagnosis, over-diagnosis, and consistent hypertension groups. Nonetheless, limitations should be acknowledged too. First, we analysed national health surveys with a standard protocol [10], which may not be equivalent to BP measurements in real life (e.g., massive screening programmes [12]). Unfortunately, massive screening programmes are not conducted routinely throughout the world, and where these occur, data are not available. Thus, population-based surveys are the only resource to expand the evidence about simplified BP screening approaches beyond a few countries. Second, because of data availability we only studied people aged 18–69 years. Recommendations derived from our work cannot be extrapolated to people ≥70 years of age. Third, this is a cross-sectional analysis. Although we documented that missed hypertension cases had slightly lower cardiovascular risk than those consistently diagnosed with hypertension, whether the missed cases went on to have worse cardiovascular outcomes than their peers who were consistently diagnosed with hypertension remains unknown. Large multi-country cohort studies are needed to strengthen this evidence [19]. Fourth, we used a standard cardiovascular risk score recommended for global use; however, this score may still have limitations. We used the cardiovascular risk score as a summary measure to characterise the overall cardiometabolic risk profile, not to make predictions about the cardiovascular outcomes in these populations. As discussed above, longitudinal studies are needed to characterise long-term cardiovascular outcomes for the simplified BP measurement approaches. Fifth, even though we pooled 60 national health surveys, all of which were nationally representative, for some countries the analysis included approximately 2,000 individuals after we applied our selection criteria. A sample size of approximately 2,000 people could be considered (rather) small, and could lead to variations in the estimates. However, we argue that these surveys were conducted using a standard and validated complex survey design in a random sample of the general population. Therefore, they provide informative results for the overall population in these countries. Given our selection criteria, the results for each country may not be representative of all people in the target population, but our results strongly characterise the patterns and profiles of the 9 simplified BP measurement strategies.

Conclusions

Simplified BP screening approaches, to maximise resources and to reach much more people, appear to be sensible, with low rates of missed cases, amongst whom the absolute cardiovascular risk appears to be slightly lower than in the population with diagnosed hypertension. The fact that there was large variation in the percentage of missed hypertension cases for the different simplified approaches suggests that a one-size-fits-all approach should not be applied to all countries. More in-country research is needed to identify the factors affecting such variation among the countries.

Analysis code (R).

(R)

Click here for additional data file.

STROBE checklist.

(DOC)

Click here for additional data file.

Code for data cleaning (R).

(R)

Click here for additional data file.

Proportion over-diagnosed by simplified approach and country.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach first blood pressure only.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure only.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach average of the first and second blood pressure measurements.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure measurement if first is ≥130/80 mm Hg.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure measurement if first is ≥135/85 mm Hg.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure measurement if first is ≥140/90 mm Hg.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure measurement if first is 130–145/80–95 mm Hg.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure measurement if first is 130–150/80–100 mm Hg.

(PDF)

Click here for additional data file.

Comparison of risk factors and 10-year absolute cardiovascular risk across classification status and by country based on the simplified approach second blood pressure measurement if first is 130–155/80–105 mm Hg.

(PDF)

Click here for additional data file.

Histogram of 10-year absolute cardiovascular risk by country.

(PDF)

Click here for additional data file.

Selection of study population.

(DOCX)

Click here for additional data file.

Countries and regions.

(CSV)

Click here for additional data file.

Proportion of over-diagnosis overall.

(CSV)

Click here for additional data file.

Proportion of over-diagnosis by region.

(CSV)

Click here for additional data file.

Description of cardiometabolic risk factors by country.

(CSV)

Click here for additional data file.

p-Values for S2 Fig.

(CSV)

Click here for additional data file.

p-Values for S3 Fig.

(CSV)

Click here for additional data file.

p-Values for S4 Fig.

(CSV)

Click here for additional data file.

p-Values for S5 Fig.

(CSV)

Click here for additional data file.

p-Values for S6 Fig.

(CSV)

Click here for additional data file.

p-Values for S7 Fig.

(CSV)

Click here for additional data file.

p-Values for S8 Fig.

(CSV)

Click here for additional data file.

p-Values for S9 Fig.

(CSV)

Click here for additional data file.

p-Values for S10 Fig.

(CSV)

Click here for additional data file.

Proportion of 10-year absolute cardiovascular risk comparing diagnosis status by country.

(CSV)

Click here for additional data file.

Standard deviations for Table 1.

(CSV)

Click here for additional data file.

Blood pressure measurement protocol by country.

(XLSX)

Click here for additional data file.

Code for computing 10-year predicted absolute cardiovascular risk according to the 2019 World Health Organization Cardiovascular Risk Charts (Stata).

(DO)

Click here for additional data file.

17 Nov 2021

Dear Dr Carrillo-Larco,

Thank you for submitting your manuscript entitled "Simplified hypertension screening: misdiagnosis, over-diagnosis and cardio-metabolic characterisation in 60 low- and middle-income countries" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by Nov 19 2021 11:59PM.

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Beryne Odeny

PLOS Medicine

25 Feb 2022

Dear Dr. Carrillo-Larco,

Thank you very much for submitting your manuscript "Simplified hypertension screening: misdiagnosis, over-diagnosis and cardio-metabolic characterisation in 60 low- and middle-income countries" (PMEDICINE-D-21-04705R1) for consideration at PLOS Medicine.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Mar 18 2022 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

1) Please include line numbers in your next draft

2) Please revise your title according to PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible and include the setting of the study. Please place the study design (e.g., “a multi-country evaluation of surveys”) in the subtitle (i.e., after a colon).

3) Abstract:

a) Please structure your abstract using the PLOS Medicine headings (Background, Methods and Findings, Conclusions). Please merge Methods and Results into one section “Methods and Findings”

b) Please ensure that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text.

c) Please include the study design, regions and setting, total number of participants, month/year during which the surveys took place

d) Please include the specific important dependent variables that are adjusted for in the analyses.

e) In the last sentence of the Abstract’s Methods and Findings section, please describe the main limitation(s) of the study's methodology.

f) Abstract Conclusions: Please address the study implications without overreaching what can be concluded from the data.

4) At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

5) Introduction: Please conclude the Introduction with a clear description of the study question or hypothesis.

6) Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section.

a) If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study, and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b) If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c) In either case, changes in the analysis-- including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

7) Please ensure that the study is reported according to the STROBE guideline, and include the completed STROBE checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)." The STROBE guideline can be found here: http://www.equator-network.org/reporting-guidelines/strobe/

8) Please provide 95% CIs and p values, where relevant, for all estimates in the text and tables.

9) Please do not report p values to 3 decimal points

10) In the Discussion section, please delete the titles “results in context” and “clinical medicine and public health implications”

11) Please define the abbreviations in Tables and Figure e.g., BMI in Table 5.

12) Please remove the “Conflict of Interests”, “Data availability,” and “Funding” statements at the end of the main text. This information is captured in the metadata obtained in the submission form

13) References:

a) Please select the PLOS Medicine reference style in your citation manager. In-text reference call outs should be presented as follows noting the absence of spaces within the square brackets, e.g., "... services [1,2]."

b) References should have no less and no more than six names before et al.

Comments from the reviewers:

Reviewer #1: This is a well-conducted study on simplified hypertension screening with regard to misdiagnosis, over-diagnosis and cardio-metabolic characterisation in 60 low- and middle-income countries. The study design, datasets, and statistical methods and analyses are mostly adequate. However, there are still a few issues needing attention, especially in presentation of the results.

1) The study is mostly descriptive and would be good to keep it descriptive to present all the facts on missing hypertentions in the 9 simplified screening approaches. The 10-year cardiovascular risk is just a risk score which itself has limitations in validation and accuracy, and they are a few others of similar kind. As all the surveys are cross-sectional studies, it would be good to focus on the observed facts rather than projections. Both Table 4 and Figure 2 are huge and not very informative and also on 10-year cardiovascular risk, suggest to move both to the supplementary information, and also tone down the claims on 10-year risks as subject to scrutiny and more work is needed.

2) Only around a couple of thousands paticipants in the survey in each country which is subject to variation and bias for the estimates. Can authors please go a bit comprehensive and critical in the limitation section to discuss the potential impact on the results of the study?

3) Table 5 on factors contributing to missed hypertension diagnosis. Age and BMI are mostly consistent but cadiovascular risk is not really, only 4 of 9 have significant results. Any interpretations and cautions to read the results?

Reviewer #2: This is a very well written, clear, and important paper.

The authors identified the share of misdiagnoses and over-diagnoses when using simplified hypertension screening approaches.

Given that hypertension screening needs to be increased globally in the fight against the epidemiological transition, more information on how screening efforts can be increased and made more efficient are highly relevant.

The authors use adequate statistical methods in answering their research question.

I have only minor comments and suggestions - and the authors may choose to ignore them.

- One sentence is a bit unclear (page 4 and 5): "We used the same predictors except for diabetes; the original model included history of diabetes whereas we included history and new diabetes cases (i.e., aware and unaware cases, the latter defined with fasting plasma glucose ≥126 mg/dL)"

It should be made clearer which predictors this is referring to.

- Perhaps the authors could add more details on how consequential simplified hypertension screenings may actually be for the CVD disease burden. This could be done by:

- including references for HTN screening being still too low in LMICs (see this study https://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(19)30955-9.pdf )

- discussing how relevant it actually is to make screening time shorter. I wonder how much time is actually being saved by taking only e.g. only one measurement (could briefly mention wait time in between measurements). As the authors have pointed out, this may be particularly relevant in massive screening programs - but how frequent are these or should even be as screening efforts are being ramped up?

Reviewer #3: This is an interesting approach to an important question. It's a big paper with a lot of analyses. I have one substantial criticism and a few smaller ones.

The substantial criticism, which the authors responsibly describe in their limitations, is that the protocol-driven hypertension assessment of STEPS has poor alignment with real-world BP practices. A fair amount of evidence shows clinic values are less reliable, wider variability, and generally don't do the three-measure assessment this study is examining. It's testing a measurement approach that I basically think doesn't exist in the world. This makes the external utility of this work unclear to me.

The smaller complaints all stem from, to my reading, them not quite succeeding at making a hard topic simple enough for the reader. There are a lot of analyses and I never quite found the "story" in their outcomes. There are 3 things that make this hard to write about. First is that diagnostic testing is a difficult field. The established terms of sensitivity and specificity are confusing and unnatural. Seemingly to avoid that they used non-standard terms like proportion missed and over-diagnosed. But these were hard to track and weren't reported consistently. The core findings - the classic 2x2 table of missed and overdiagnosed relative to true and non hypertensive - was clear in the abstract, but not elsewhere.

Second, cross-national comparisons are hard to write because it's hard to see the story. A table of 60 countries is very hard to get a feel for or read, especially because it's unlikely that 2 countries of the same region actually have different phenomena. The manuscript dealt with this difficulty 3 different ways. They had a few tables with 60 countries, a few divided by continent, and some organized internationally with multilevel modeling (Table 5). I can see the urge for all of these, but it didn't help me understand what was going on. What do we learn from each? Do we think it's a national, regional, or single international phenomenon?

This became especially difficult in the figures, which don't really work for me. They showed statistical significance, which is a feature of sample size, effect size, and random chance. In this case we know the effect is real - that the new measures do not precisely match the old - so there's no question we'd reach significance with enough sample. But I don't know the sample size of each of the 60 rows and this obscures the effect size, which is more interesting. I generally didn't find the story behind the figures that compelling compared to the general issue of learning how reliable the techniques are. (Though the fact, which appeared in a few ways, that the errors were larger in those with high risk and history of CVD is both concerning and interesting.)

The third difficulty was just that nine is a lot of comparisons. This made some tables feel overwhelming. But to address that, they didn't include all 9 at all spots in the paper, like the figures, even though we'd just seen them in a table.

I think if this paper is resubmitted, I would recommend really deciding what the story is and telling that. Then, putting the details in the supplement and acknowledging in the paper that what is presented is post-hoc. I would put the missed and overdiagnosis in the same tables. I would report sensitivity and specificity somewhere. I would recommend grouping the countries in a way that feels meaningful and only putting all 60 in the main paper once. (Put more in an appendix.)

I have a few questions:

1. Table 5 seems to show that higher risk people are are likely to have missed HTN. But Table 4 seems to show that missed hypertension have a risk of CVD that is roughly half the risk of those with consistent hypertension. How?

2. In table 5 the HR of CV risk is ~3 in all examples except one where it's 0.89. Is that right?

3. I would benefit by having them talk through the implications of the random effects findings in table 5.

4.Table 2 could use something like "by country" somewhere. I struggled to understand what the SD was the SD of, though I now see it's of the results of each country.

5. I would probably merge tables 2 and 3, remove the min and max, and add columns for overdiagnosis into that table.

Reviewer #4: Blood pressure measurement is key in the screening and management of high blood pressure. In the most comprehensive scenario patients would need multiple BP measurements

on separate occasions, and in the most pragmatic approach would require three BP measurements and taking the average of the last two which is standard approach. Three BP measurement may however be challenging in resource-constraint settings because of shortage health care workforce. Therefore, it will be necessary to have fewer BP measurements that would have values as close as possible to taking the average of the last two measurements which is currently the gold standard. Although attempts have been made to find simplified BP screening approaches, these simplified approaches were tested in three countries only and BP may vary between countries. The authors then decided to analyse national surveys in 60 low- and middle-income countries giving a large dataset. The authors further went ahead to relate the different measurement to their effect on cardiovascular risk score.

This paper is well written and a stimulation at looking for ways that can make blood pressure measurement especially during mass screening programmes to be more pragmatic.

Apart from country to country variations as stated by the authors is there any regional differences to the variation? Although one-size-fits will be applicable to different countries it will be nice as a more pragmatic approach for the authors based on their study to state the 2 or 3 methods out of the nine tested that they will recommend for countries to be tested.

Reviewer #5: Authors considered only BP measurement for hypertension diagnosis but currently having BP lowering drugs, earlier suggestion by a practitioner about hypertension as well as life style modification for hypertension management are part of hypertension definition which are ignored by authors.

Reference 15 seems incorrect or broken.

Any attachments provided with reviews can be seen via the following link:

[LINK]

2 Mar 2022

Submitted filename: Rebuttal 20220301.docx

Click here for additional data file.

22 Mar 2022

Dear Dr. Carrillo-Larco,

Thank you very much for re-submitting your manuscript "Simplified hypertension screening: a multi-country evaluation of surveys" (PMEDICINE-D-21-04705R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by two reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Mar 29 2022 11:59PM.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

1. Please consider the suggested title, e.g., “Simplified hypertension screening methods across 60 countries: An observational study,” or similar.

2. Thank you for providing your STROBE checklist. Please replace the page numbers with paragraph numbers per section (e.g. "Methods, paragraph 1"), since the page numbers of the final published paper may be different from the page numbers in the current manuscript.

3. Please include p-values in tables where appropriate e.g., Table 4

4. Please replace the term “predictor” with “independent variable.” e.g., table 4

5. References – please include access dates for references with a weblink e.g. ref #6

Comments from Reviewers:

Reviewer #1: Many thanks authors for their great effort to improve the manuscript. The authors have addressed my comments very well. I am satisfied with the response and revision. No further issues needing attention.

Reviewer #5: Congratulations for great work

Any attachments provided with reviews can be seen via the following link:

[LINK]

24 Mar 2022

Dear Dr Carrillo-Larco,

On behalf of my colleagues and the Academic Editor, Dr. Joshua Z Willey, I am pleased to inform you that we have agreed to publish your manuscript "Simplified hypertension screening methods across 60 countries: An observational study" (PMEDICINE-D-21-04705R3) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Beryne Odeny

PLOS Medicine