Advertisement

Prevalence and risk factors for atrial fibrillation in a semi-rural sub-Saharan African population: The hEart oF ethiopia: Focus on Atrial Fibrillation (TEFF-AF) Study

      Background

      There is a scarcity of reported data on the prevalence of atrial fibrillation (AF) in sub-Saharan Africa.

      Objectives

      To undertake AF screening in semi-rural Ethiopia.

      Methods

      The TEFF-AF (The hEart oF Ethiopia: Focus on Atrial Fibrillation) study conducted AF screening using a single-lead electrocardiogram device (KardiaMobile) on willing community participants at the Soddo Christian Hospital, Ethiopia. Participants’ clinical parameters and medical history were obtained to characterize their risk factor profile, including calculation of CHARGE-AF (Cohorts for Heart and Aging Research in Genomic Epidemiology Atrial Fibrillation) score.

      Results

      A total of 3000 Ethiopians (median 31 [interquartile range 25–41] years of age; 65% men) were screened. The participants were generally well educated, from the local region and with a low burden of cardiovascular risk factors. A total of 50 participants had a CHARGE-AF score (5-year AF risk) of ≥2%. AF was detected in 13 (0.43%) individuals (median 50 [interquartile range 36–60] years of age; n = 7 men). The prevalence among participants over 40 years of age was 1% (n = 9 of 930). AF prevalence was higher for older age groups, with ≥70 years of age reaching 6.67% (n = 3 of 45). Population prevalence was estimated to be 234 (95% confidence interval 7–460) per 10,000 persons for ≥60 years of age. Four (31%) of the 13 participants with AF had a CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65-74 years, sex category) score of ≥2, and others likely had rheumatic valvular AF, but only 2 of the 13 participants with AF were on oral anticoagulation therapy.

      Conclusion

      In this semi-rural Ethiopian community of relatively younger participants, AF prevalence was found to be low but increased with increasing age. Mobile single-lead electrocardiogram technology can be used effectively for AF screening in low-resource settings.

      Keywords

      Key Findings
      • Overall AF prevalence was 0.43% in a semi-rural Ethiopian community of 3,000 generally younger (median 31 years old) participants.
      • AF prevalence was higher with increasing age and in the presence of known structural heart disease.
      • Hypertension was the most common risk factor for AF followed by diabetes and valvular heart disease.
      • 31% of the participants screened with AF in this study had high CHA2DS2-VASc score (≥3 for females, ≥2 for males) and were not on oral anticoagulation therapy.

      Introduction

      Atrial fibrillation (AF) is a growing public health problem worldwide. The Global Burden of Disease Study estimated that the worldwide age-adjusted prevalence of AF in 2010 was approximately 0.5%, representing a total of 33.5 million individuals.
      • Chugh S.S.
      • Havmoeller R.
      • Narayanan K.
      • et al.
      Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study.
      In sub-Saharan Africa, the leading cardiovascular cause of death and disability in 2010 was stroke, and the largest relative increases in cardiovascular disease burden between 1990 and 2010 were in AF and peripheral arterial disease.
      • Chugh S.S.
      • Havmoeller R.
      • Narayanan K.
      • et al.
      Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study.
      Furthermore, cardiovascular deaths occur at younger ages in sub-Saharan Africa as compared with the rest of the world. Individuals with AF in Africa have higher mortality rates,
      • Healey J.S.
      • Oldgren J.
      • Ezekowitz M.
      • et al.
      Occurrence of death and stroke in patients in 47 countries 1 year after presenting with atrial fibrillation: a cohort study.
      although due largely to poor health care access and suboptimal therapy. The Global Burden of Disease Study also highlighted the crucial need for more data from population surveillance studies in sub-Saharan Africa due to the paucity of data from this region.
      Several factors may contribute to the changing prevalence of AF in developing nations. The burden of communicable disease persists and rheumatic heart disease is associated with higher AF prevalence in younger populations.
      • Stambler B.S.
      • Ngunga L.M.
      Atrial fibrillation in sub-Saharan Africa: epidemiology, unmet needs, and treatment options.
      Additionally, the epidemiological transition in sub-Saharan African regions with gradual adoption of Western lifestyle is likely increasing the impact of modifiable risk factors on AF prevalence. There is significant variation in the prevalence of AF reported across different countries in Africa, and estimates may be skewed by poor health-seeking behaviors and limited access to diagnostic equipment.
      • Adedinsewo D.
      • Omole O.
      • Oluleye O.
      • Ajuyah I.
      • Kusumoto F.
      Arrhythmia care in Africa.
      Better understanding of the cardiovascular risk factor profile in these regions can help guide preventive and management strategies. Given the relative scarcity of AF epidemiology data from sub-Saharan African regions, we undertook the TEFF-AF (The hEart oF Ethiopia: Focus on Atrial Fibrillation) screening study to characterize the prevalence of AF and associated risk factor profile in semi-rural south-central Ethiopia.

      Methods

      The TEFF-AF study (ACTRN12619001107112) was undertaken from August through December 2019, on the campus of the Soddo Christian Hospital (SCH), which provided institutional research ethics approval. This study complied with the ethical principles of the 2013 Declaration of Helsinki. SCH is a major trauma center situated in the town of Soddo, with a population of around 200,000 individuals in south-central Ethiopia. Eligibility was limited to ambulant adults 18 years of age and above who were able to provide informed consent. Signage in the Amharic language was erected at SCH to invite willing individuals from the community and visitors to the hospital to participate in the screening. Inpatients at SCH were not included in this study. A team of 5 nursing and research support staff from the SCH performed the screening following specialized training on the use of the KardiaMobile (KM) single-lead electrocardiogram (ECG) device (AliveCor, Mountain View, CA) paired to an iPhone application (version 5.7.4, KardiaAI: 1.1.7). Training included tutoring on acquiring the best quality single-lead ECG tracing with the KM device and subsequent hands-on practice.
      An online customized REDCap (Research Electronic Data Capture) database was utilized.
      • Harris P.A.
      • Taylor R.
      • Minor B.L.
      • et al.
      The REDCap consortium: building an international community of software platform partners.
      Baseline demographic and clinical parameters were obtained by a brief structured questionnaire to characterize the cardiovascular risk profile of participating individuals. The CHARGE-AF (Cohorts for Heart and Aging Research in Genomic Epidemiology Atrial Fibrillation) risk model adequately predicts 5-year AF risk among Africans in the United States and Europe (using variables of age, race, height, weight, systolic and diastolic blood pressure, current smoking, use of antihypertensive medication, diabetes, and history of myocardial infarction and heart failure).
      • Alonso A.
      • Krijthe B.P.
      • Aspelund T.
      • et al.
      Simple risk model predicts incidence of atrial fibrillation in a racially and geographically diverse population: the CHARGE-AF consortium.
      This score was calculated for each TEFF-AF study participant. Stroke risk was calculated using CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65-74 years, sex category) score for participants found to have AF. A multistage protocol for screening was undertaken, which has previously been described.
      • Pitman B.M.
      • Chew S.H.
      • Wong C.X.
      • et al.
      Performance of a mobile single-lead electrocardiogram technology for atrial fibrillation screening in a semirural african population: insights from "The Heart of Ethiopia: Focus on Atrial Fibrillation" (TEFF-AF) study.
      Height and weight measurements were obtained to calculate body mass index (BMI). Blood pressure measurements were undertaken in the seated position (Omron Intellisense T5 automatic monitor; Omron Corporation, Kyoto, Japan), which was repeated when the initial systolic reading exceeded 130 mm Hg. A 30-second single-lead ECG acquisition using the KM device was then obtained from the participant.

      ECG acquisition and AF determination

      The KM device records a bipolar lead I ECG tracing when 2 or 3 fingers from each hand of the user are placed in contact with the 2 electrodes. Participants were instructed to relax their arms and hands to reduce noise and artefact. A 30-second single-lead ECG recording can be viewed in real time on the smartphone application and saved as a PDF file. The noise-filtered trace and computer-averaged complex on the smartphone application are then subjected to an automated algorithm for arrhythmia diagnosis using the 2 criteria of P-wave absence and R-R interval irregularity.
      • Lau J.K.
      • Lowres N.
      • Neubeck L.
      • et al.
      iPhone ECG application for community screening to detect silent atrial fibrillation: a novel technology to prevent stroke.
      The outcome of the automated algorithm assessment of rhythm dictated the need for repeat KM tracing or a 12-lead ECG. The screening protocol is deemed complete if the rhythm was assessed as normal by the automated algorithm. A 12-lead ECG was indicated if the rhythm was assessed as possible AF. A repeat KM tracing was required if the rhythm was assessed as bradycardia, tachycardia, unclassified, unreadable, or too short. If the same results or possible AF were obtained after repeat KM tracing, a 12-lead ECG will then be indicated.
      The 30-second KM traces obtained by participants were downloaded as PDFs for manual analysis by 2 cardiologists independently (investigators C.X.W. and S.I.) to confirm rhythm diagnosis. The cardiologists also assessed diagnostic limitations for traces categorized automatically as bradycardia, tachycardia, unclassified, unreadable, or too short into the following: artifact, ectopy, bradycardia, tachycardia, or insufficient sample duration. All 12-lead ECGs were standardized with scale of 1 mV:10 mm and paper speed of 25 mm/s. They were independently adjudicated by 2 physicians (investigators S.-H.C. and D.H.L). Participants in AF were offered transthoracic echocardiogram at SCH to evaluate the presence of any structural heart disease and referred for follow-up by SCH physician.

      Statistical analysis

      Continuous data are expressed as mean ±SD or median (interquartile range [IQR]) appropriate to data distribution assessed by Shapiro-Wilk test. Binary data are presented as percentage (with numerator and denominator in brackets). Categorical data were analyzed using the chi-square or Fisher exact test. The sensitivity and specificity for the ability of the KM to produce a rhythm decision against the cardiologists’ interpretation were calculated. All statistics were performed in SPSS (version 26; IBM, Armonk, NY), with statistical significance set at P < .05.

      Results

      A total of 3000 participants were recruited. The baseline clinical parameters of the participants are shown in Table 1. The median age was 31 (IQR 25–41) years (35% women; median BMI 23.0 [IQR 20.5–26.4] kg/m2), with 31% being 40 years of age or older. The vast majority (94%) of the participants were from the regional state of Southern Nations, Nationalities, and People's Region, where SCH is located. Seventy-eight percent (n = 2352) had at least secondary-level education, and 44% (n = 1320) reported salaried employment.
      Table 1Baseline clinical characteristics (n = 3000)
      Demographic and clinical information (n = 3000)Data
      Age, y31 (25–41)
       <30 y1224 (41)
       30–39 y846 (28)
       40–49 y472 (16)
       50–59 y287 (10)
       60–69 y126 (4)
       70+ y45 (2)
      Sex
       Male1975 (65)
       Female1025 (35)
      Home region
       SNNPR2810 (94)
       Oromia61 (2)
       Amhara17 (1)
       Other regions (including, Somalia, B-Gumuz, Addis Ababa, Harar)47 (2)
       Unspecified65 (2)
      Religion
       Orthodox786 (26)
       Protestant2015 (67)
       Muslim129 (4)
       Other religion or no religion70 (2)
      Education
       Illiterate160 (5)
       Primary level school462 (15)
       Secondary level school1028 (34)
       Certificate, diploma, or above1324 (44)
       Unspecified26 (1)
      Occupation
       Unemployed84 (3)
       Employed (worker)484 (16)
       Employed (professional)836 (28)
       Self-employed737 (25)
       Householder/housewife251 (8)
       Retired22 (1)
       Student472 (16)
       Other114 (4)
      Clinical
       BMI, kg/m223.0 (20.5–26.4)
      Underweight (<18.5 kg/m2)223 (7)
      Normal (18.5–24.9 kg/m2)1749 (58)
      Overweight (25–29.9 kg/m2)770 (26)
      Obese (≥30 kg/m2)258 (9)
       Systolic BP, mm Hg124 (114–135)
      Hypertensive (systolic BP >140 mm Hg)533 (18)
      Values are median (interquartile range) or n (%).
      BMI = body mass index; BP = blood pressure; SNNPR = Southern Nations, Nationalities, and People's Region.

      Prevalence of AF

      There were a total of 13 participants (n = 7 men) determined to have AF, equating to an overall prevalence of 0.43% (Figure 1A). The prevalence of AF increased exponentially with increasing age, from 0.18% in those under 30 years of age to 2.3% in those above 60 years of age. The AF prevalence for participants ≥70 years ofage was 6.67% (n = 3 of 45). The estimation of AF prevalence for each age group is shown in Figure 1B, with 234 (95% confidence interval 7–460) per 10,000 persons among those 60 years of age and above. The clinical parameters of the participants with AF are shown in Table 2. History of known structural heart disease was reported by 6 of these participants, and 4 participants had a systolic blood pressure above 130 mm Hg during screening. The median heart rate of those in AF was 94 (IQR 85–100) beats/min, including 1 individual at 148 beats/min. Transthoracic echocardiogram was obtained for 7 of these participants, demonstrating presence of underlying valvular pathology in all of them.
      Figure thumbnail gr1
      Figure 1Atrial fibrillation (AF) prevalence. A: Frequency of AF detected by screening for men and women by age group and B: estimated AF prevalence per age group. CI = confidence interval.
      Table 2Participants with atrial fibrillation
      Age (y); sexBMI (kg/m2)Medical historyMedicationsBP (mm Hg); heart rate (beats/min)CHA2DS2-VASC ScoreEchocardiogram
      20; M20.2CHF, heart diseaseEnalapril, warfarin100/80; 1281Dilated LV with systolic dysfunction, dilated LA, MR, and TR
      20; M16.3CHF, heart disease (valvular)Enalapril, metoprolol, warfarin, frusemide120/70; 1001Dilated LV with systolic dysfunction, dilated LA, AR/MR/TR
      30; M23.9Heart disease (valvular)Monthly benzylpenicillin122/90; 880N/A
      36; F23.8HTN, DMNone117/86; 993Normal size LV and low-normal systolic function, dilated LA, MR/TR, pulmonary hypertension, RV dilated with poor systolic function
      40; F26.3MalariaNone132/92; 941N/A
      48; F19.4None reportedNone127/86; 831N/A
      50; M18.8CHF, Heart disease(valvular)Enalapril, phenytoin, statin, aspirin120/70; 1001Normal size LV, grossly dilated LA, severe MS, severe MR, and severe TR
      50; F24.6Thyroid diseaseNone145/85; 801N/A
      58; M19.8None reportedNone100/64; 850N/A
      60; M17.0None reportedNone116/83; 690N/A
      70; F21.3HTN, DM, CHFFrusemide, aspirin, enalapril, digoxin, spironolactone145/83; 1085Normal size LV, dilated LA, MR, and TR
      75; M19.1CHF, Heart disease (valvular)Enalapril, frusemide, spironolactone116/88; 1483Valvular regurgitation (AR/MR/TR)
      80; F18.2HTN, currently has pneumoniaNone150/80; 654Normal biventricular size and function; moderate AR/AS and mild MS
      AR = aortic regurgitation; AS = aortic stenosis; BMI = body mass index ; BP = blood pressure; CHA2DS2-VASc = congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65-74 years, sex category; CHF = congestive heart failure; F = female; DM = diabetes mellitus; HTN = hypertension; LA = left atrium; LV = left venticle; M = male; MR = mitral regurgitation; MS = mitral stenosis; N/A = not available; RV = right ventricle; TR = tricuspid regurgitation.

      Associated comorbidities in those with and without AF

      Participants with AF were more likely to be older, with a higher proportion with known heart failure, with valvular heart disease, and on regular medications (Table 3). The most prevalent reported risk factor was hypertension (5.2%), although 17.8% (n = 533 of 3000) of the subjects had systolic blood pressure >140 mm Hg during screening. Malaria and typhoid (10% and 14%, respectively) were common among participants but not for those with AF. A total of 50 participants had a CHARGE-AF score (5-year AF risk) of ≥2%. The CHARGE-AF score was on average higher for those with AF than for those without AF (0.84% vs 0.22%; P = .029), although only 3 (23%) participants with AF had a CHARGE-AF score of ≥2% (5-year risk of AF) (Figure 2A). Four (31%) of the 13 participants with AF had a CHA2DS2-VASc score of 2 or greater (Figure 2B), consisting of 3 women with CHA2DS2-VASc score ≥3 and 1 man with CHA2DS2-VASc score ≥2, while none of these participants were on oral anticoagulation therapy. Only 2 other participants with AF were on oral anticoagulation therapy, and both were men with congestive heart failure associated with mitral or aortic regurgitation (Table 2).
      Table 3Participant clinical parameters, self-reported medical history, and risk factors
      Participants without AF (n = 2987)Participants with AF (n = 13)P value
      age, y31 (25–41)50 (36–60).005
      age ≥65 y76 (3)3 (23).18
      BMI, kg/m223 (21–26)20 (19–24).007
      Measured systolic blood pressure ≥140 mm Hg530 (18)3 (23).71
      Hypertension153 (5)3 (23).27
      Diabetes mellitus66 (2)2 (15).41
      Heart failure19 (1)7 (54)<.001
      Stroke3 (<1)0 --
      Valvular heart disease7 (<1)4 (31).046
      Renal failure5 (<1)0
      Chronic lung disease16 (1)0
      Obstructive sleep apnoea2 (<1)0
      Thyroid disease21 (1)1 (8).79
      Smoker
       Current7 (<1)0
       previous19 (1)1 (8).78
      Khat chewing
       Current11 (<1)0
       previous21 (1)0
      Alcohol
       Current7 (<1)0
       previous17 (1)0
      Malaria409 (14)0
      Typhoid308 (10)0
      any infective disease
       Current50 (2)2 (15).38
       previous233 (8)2 (15).80
      Currently taking any medication128 (4)7 (54).001
      Values are median (interquartile range) or n (%).
      AF = atrial fibrillation.
      Figure thumbnail gr2
      Figure 2CHARGE-AF (Cohorts for Heart and Aging Research in Genomic Epidemiology Atrial Fibrillation) and CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65-74 years, sex category) scores. A: Calculated CHARGE-AF score for 5-year risk of AF and B: calculated CHA2DS2-VASc score for AF participants. TEFF-AF = The hEart oF Ethiopia: Focus on Atrial Fibrillation.

      Outcomes from KM automated algorithm

      Of the initial single-lead KM tracings performed on the 3000 participants (Figure 3A), the KM algorithm was unable to provide a rhythm decision in 18% (n = 549) due to being categorized as unclassified (8.6%, n = 258), tachycardia (7.2%, n = 215), unreadable (2.1%, n = 63), too short (0.1%, n = 3), and bradycardia (0.3%, n = 10). A repeat KM tracing was obtained in 70% (n = 383 of 549) of the participants who did not have an initial rhythm decision, which achieved a rhythm decision in 207 additional subjects, with an overall rhythm decision in 89% (n = 2658) (Figure 3B and 3C). The remaining repeat KM tracings showed tachycardia’ (n = 79), unclassified (n = 75), unreadable (n = 12), and bradycardia (n = 10). Manual assessment of the KM tracings by cardiologists provided a greater rhythm decision yield than the KM algorithm. Only 2% of KM traces were considered nondiagnostic by manual adjudication. Overall, the AliveCor KM had a sensitivity of 81.3% and specificity of 96.5%, and although the negative predictive value of 99.9% was high, the positive predictive value was only 12.4%.
      Figure thumbnail gr3
      Figure 3Participants’ rhythm assessment by KardiaMobile (KM) automated algorithm. A: KM algorithm outcome on first attempt. B: KM algorithm outcome on second attempt. C: KM algorithm outcome on third attempt. AF = atrial fibrillation.

      Twelve-lead ECG analysis

      In total, there were 456 (15%) participants who met protocol criteria for a 12-lead ECG, but only 181 ECGs were obtained (Figure 4). This was largely due to participants not wanting to wait for the 12-lead ECG to be performed in the SCH emergency room, or a KM outcome of tachycardia that the screening team assessed as sinus tachycardia not requiring a 12-lead ECG. The majority (n = 160 of 181) of 12-lead ECGs showed sinus rhythm. Twelve-lead ECG was performed in 65% (n = 74) of the 114 possible AF patients. However, manual cardiologists’ assessment of the remaining 40 possible AF KM traces without 12-lead ECGs only found 2 AF tracings. There were no 12-lead ECG diagnosis of AF not having a KM outcome of possible AF. A repeat KM attempt without a rhythm decision indicated the need for a 12-lead ECG for 342 participants. However, a 12-lead ECG was obtained for only 74% (n = 17 of 23) of unreadable, 64% (n = 67 of 104) of unclassified, 33% (n = 4 of 12) of bradycardia, and 9% (n = 19 of 203) of tachycardia assessments. Of these 107 ECGs, the majority were adjudicated as sinus rhythm or sinus arrhythmia, and the remainder were found to have intraventricular conduction delay, pre-excitation, multiple ectopics, supraventricular tachycardia, complete heart block, or junctional rhythm (Figure 4).
      Figure thumbnail gr4
      Figure 4Diagnosis by 12-lead electrocardiogram (ECG). Diagnostic yield of 12-lead ECGs performed for each final KardiaMobile outcome. AF = atrial fibrillation; IVCD = intraventricular conduction delay.,

      Participant experience and protocol compliance

      During the screening protocol, the majority of participants (84%, n = 2531) only performed 1 KM ECG trace without a repeat or a 12-lead ECG. There were additionally 10% (n = 288) who completed 2 KM attempts but did not have a 12-lead ECG. A 12-lead ECG was performed in addition to the KM screening for the remaining 6% (n = 181). Overall, 90% (n = 2694) of participants completed the screening protocol as defined. Noncompliancy was due to not having performed a required repeat KM in 5.6% (n = 168) and not having performed a required 12-lead ECG in 3.6% (n = 107). Additionally, there were 31 participants who did not undergo repeat KM ECG acquisition but had a 12-lead ECG performed instead. There were 21 participants who provided an additional KM trace beyond the protocol requirement who did not have a 12-lead ECG performed as required after a KM algorithm result of possible AF. There were only 2 occasions when this repeat KM attempt again returned a result of possible AF.

      Discussion

      This study is the largest AF screening study from semi-rural Ethiopia. First, we found overall AF prevalence of 0.43% in 3000 generally younger (median 31 years of age) and well-educated participants. Second, AF prevalence was higher with increasing age and in the presence of known structural heart disease. Third, hypertension was the most common risk factor for AF, followed by diabetes and valvular heart disease. Fourth, the KM single-lead ECG technology was suitable for AF screening with the automated algorithm able to detect 85% as normal sinus rhythm. Additional 12-lead ECG and manual assessment was needed for rhythm assessment in the remainder. Last, 31% of the participants screened with AF in this study had CHA2DS2-VASc score of more than 2 and were not on oral anticoagulation therapy. Our findings have important public health implications, given the scarcity of such reported data from this community.
      • Freedman B.
      • Camm J.
      • Calkins H.
      • et al.
      Screening for atrial fibrillation: a report of the AF-SCREEN international collaboration.
      In a prospective global registry of >15,000 AF patients, it was observed that annual AF mortality in Africa was twice that as compared with North America, Western Europe, and Australia, and the rate of strokes was highest in African patients with AF.
      • Healey J.S.
      • Oldgren J.
      • Ezekowitz M.
      • et al.
      Occurrence of death and stroke in patients in 47 countries 1 year after presenting with atrial fibrillation: a cohort study.
      Furthermore, as life expectancy in Ethiopia has increased by approximately 22 additional years over the last 2 decades (47 years in 1990 to 69 years by 2019),
      • Misganaw A.
      • Naghavi M.
      • Walker A.
      • et al.
      Progress in health among regions of Ethiopia, 1990–2019: a subnational country analysis for the Global Burden of Disease Study 2019.
      the impact of aging and chronic diseases such as AF will become more significant.

      Ascertaining AF prevalence from screening

      It is known that the incidence of detecting previously undiagnosed AF is dependent on the underlying population risk of AF, the intensity or duration of screening, and the method used to detect AF.
      • Freedman B.
      • Camm J.
      • Calkins H.
      • et al.
      Screening for atrial fibrillation: a report of the AF-SCREEN international collaboration.
      In a nearby Southwest town from our screening venue, a community-based cross-sectional study with 12-lead ECG in 634 adults (mean 63 years of age) reported an AF prevalence of 4.3%.
      • Tegene E.
      • Tadesse I.
      • Markos Y.
      • Gobena T.
      Prevalence and risk factors for atrial fibrillation and its anticoagulant requirement in adults aged >/=40 in Jimma Town, Southwest Ethiopia: A community based cross-sectional study.
      Consistent with the results of our screening, the prevalence of AF rose steeply with advancing age. They reported a prevalence of 6.9% (n = 4 of 58) in those above 80 years of age, which is comparable to the 6.67% prevalence in those above 70 years of age in this current study. Although our 30-second single-lead ECG screening found an overall low prevalence of AF in a relatively young cohort, it remains in the ballpark of the estimated worldwide age-adjusted prevalence of ∼0.5% and the 2010 Global Burden of Disease Study estimation of 0.44%–0.66% in the sub-Saharan African region.
      • Chugh S.S.
      • Havmoeller R.
      • Narayanan K.
      • et al.
      Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study.
      However, these estimates may be skewed by the effects of sampling, with known variability in health care infrastructure and services among sub-Saharan African countries and poor health-seeking behavior.
      • Adedinsewo D.
      • Omole O.
      • Oluleye O.
      • Ajuyah I.
      • Kusumoto F.
      Arrhythmia care in Africa.
      Notably, AF patients in sub-Saharan Africa tend to be younger and with higher mortality rates due to poor health care access and suboptimal therapy,
      • Stambler B.S.
      • Ngunga L.M.
      Atrial fibrillation in sub-Saharan Africa: epidemiology, unmet needs, and treatment options.
      and screening may help to identify individuals who can benefit from risk factor modification and medical therapy to reduce the burden of disease and associated morbidities.

      Impact of digital technology for AF screening in the low-resource settings

      AF screening by palpation is a simple technique that is useful in low-resource settings, but portable digital technology, such as handheld single-lead ECGs or plethysmography devices, are becoming more readily available and have shown greater yield.
      • Giebel G.D.
      • Gissel C.
      Accuracy of mHealth devices for atrial fibrillation screening: systematic review.
      We used the KM device, which has previously been shown to have high sensitivity and specificity for AF detection in validation study.
      • Lau J.K.
      • Lowres N.
      • Neubeck L.
      • et al.
      iPhone ECG application for community screening to detect silent atrial fibrillation: a novel technology to prevent stroke.
      The utility of this technology for mass screening in low-resource communities has previously been demonstrated.
      • Soni A.
      • Karna S.
      • Patel H.
      • et al.
      Study protocol for Smartphone Monitoring for Atrial fibrillation in Real-Time in India (SMART-India): a community-based screening and referral programme.
      However, we reported a limitation of this technology with the inability of the device algorithm to provide a rhythm decision for ∼13% of participants even with repeat tracing.
      • Pitman B.M.
      • Chew S.H.
      • Wong C.X.
      • et al.
      Performance of a mobile single-lead electrocardiogram technology for atrial fibrillation screening in a semirural african population: insights from "The Heart of Ethiopia: Focus on Atrial Fibrillation" (TEFF-AF) study.
      This limitation can be minimized through increased user familiarity and careful execution of the acquisition technique to obtain better-quality ECG signals. One advantage of ECG tracing over plethysmography in low-resource settings is the option of remote manual adjudication by a clinician to enhance diagnostic yield and reduced dependence on automated algorithm performance.
      • Halcox J.P.J.
      • Wareham K.
      • Cardew A.
      • et al.
      Assessment of remote heart rhythm sampling using the AliveCor heart monitor to screen for atrial fibrillation: the REHEARSE-AF study.
      ,
      • William A.D.
      • Kanbour M.
      • Callahan T.
      • et al.
      Assessing the accuracy of an automated atrial fibrillation detection algorithm using smartphone technology: the iREAD study.
      Our study protocol included repeated tracing attempts, along with select use of 12-lead ECG, to optimize yield from this single-occasion screening strategy. Although mobile digital technology provides the opportunity for continuous monitoring of large cohorts as seen in the Apple Heart and Huawei studies, the affordability of such in low-resource settings and the low sensitivity of AF detection would reduce its utility.
      • Turakhia M.P.
      • Desai M.
      • Hedlin H.
      • et al.
      Rationale and design of a large-scale, app-based study to identify cardiac arrhythmias using a smartwatch: the Apple Heart Study.
      ,
      • Guo Y.
      • Wang H.
      • Zhang H.
      • et al.
      Mobile photoplethysmographic technology to detect atrial fibrillation.

      Risk factors for AF in sub-Saharan Africa

      The risk factors for AF in sub-Saharan Africa are likely to be altered by the high prevalence of rheumatic heart disease, which is also known to be the most common acquired cardiovascular disease in young individuals under 25 years of age.
      • Lau D.H.
      • Nattel S.
      • Kalman J.M.
      • Sanders P.
      Modifiable risk factors and atrial fibrillation.
      ,
      • Yuyun M.F.
      • Sliwa K.
      • Kengne A.P.
      • Mocumbi A.O.
      • Bukhman G.
      Cardiovascular diseases in sub-Saharan Africa compared with high-income countries: an epidemiological perspective.
      Unfortunately, not much is known about the burden of AF related to rheumatic heart disease in this region.
      • Linz D.
      • Sanders P.
      • Pitman B.
      • Dobrev D.
      • Lau D.H.
      Atrial fibrillation in sub-Saharan Africa: The knowns and unknowns?.
      A recent meta-analysis has shown the global prevalence of AF in rheumatic heart disease to be 32.8% (range, 4.3%–79.9%).
      • Noubiap J.J.
      • Nyaga U.
      • Ndoadoumgue A.
      • Nkeck J.
      • Ngouo A.
      • Bigna J.J.
      Meta-analysis of the incidence, prevalence, and correlates of atrial fibrillation in rheumatic heart disease.
      In the Ethiopian capital of Addis Ababa, a retrospective chart review of 500 adult cardiology patients with rheumatic heart disease found that 46.8% had AF.
      • Yadeta D.
      • Semeredin N.
      • Mekonnen G.
      Prevalence and predictors of atrial fibrillation and its embolic complications in patients with rheumatic heart disease at Tikur Anbessa Specialized Hospital, Addis Ababa, Ethiopia.
      Indeed, valvular heart disease was more common in those with AF in our study, as confirmed with echocardiographic imaging, including the 2 on warfarin, and 1 on monthly benzylpenicillin. It is likely that the burden of valvular heart disease was higher given that 6 of the participants with AF did not undertake an echocardiogram. However, many parts of sub-Saharan Africa are undergoing epidemiological transitions, with gradual adoption of Western lifestyle leading to development of new AF risk factors including hypertension, dyslipidemia, diabetes, and obesity.
      • Tibazarwa K.
      • Ntyintyane L.
      • Sliwa K.
      • et al.
      A time bomb of cardiovascular risk factors in South Africa: results from the Heart of Soweto Study "Heart Awareness Days".
      Others have recently reported smoking, hypertension, and increased BMI as main risk factors associated with AF in Southwest Ethiopia.
      • Tegene E.
      • Tadesse I.
      • Markos Y.
      • Gobena T.
      Prevalence and risk factors for atrial fibrillation and its anticoagulant requirement in adults aged >/=40 in Jimma Town, Southwest Ethiopia: A community based cross-sectional study.
      Our data show that hypertension was the most common risk factor for AF, followed by diabetes and valvular heart disease in a younger cohort in semi-rural Ethiopia. Although no studies have demonstrated that AF screening reduces mortality or incidence of thromboembolic complications,
      • Virani S.S.
      • Alonso A.
      • Benjamin E.J.
      • et al.
      Heart Disease and Stroke Statistics—2020 Update: a report from the American Heart Association.
      we are able to characterize the risk factor profile of this semi-rural Ethiopian cohort and detect high-risk individuals with AF who are not on anticoagulants and may benefit from long-term oral anticoagulant therapy. More work is necessary to characterize the burden of AF and the associated risk factor profile including rheumatic heart disease in sub-Saharan Africa, where health care resources are scarce and more targeted public health and lifestyle measures would be valuable to curtail the expected significant increase in AF over the coming decades.
      • Stambler B.S.
      • Ngunga L.M.
      Atrial fibrillation in sub-Saharan Africa: epidemiology, unmet needs, and treatment options.
      ,
      • Lau D.H.
      • Nattel S.
      • Kalman J.M.
      • Sanders P.
      Modifiable risk factors and atrial fibrillation.
      ,
      • Linz D.
      • Sanders P.
      • Pitman B.
      • Dobrev D.
      • Lau D.H.
      Atrial fibrillation in sub-Saharan Africa: The knowns and unknowns?.
      ,
      • Hendriks J.M.
      • Gallagher C.
      • Middeldorp M.E.
      • Lau D.H.
      • Sanders P.
      Risk factor management and atrial fibrillation.

      Clinical Implications

      Our study has important clinical implications for AF screening and highlights opportunities for future research especially in the younger population with underlying structural heart disease as well as more urbanized communities that are undergoing epidemiological transition in sub-Saharan Africa. We have shown that portable digital technology can be utilized for AF screening programs in low-resource settings. Employing a strategic protocol when utilizing such technology for AF screening will ensure optimized diagnostic yield and reduce dependence on physician input to achieve screening accuracy.

      Study Limitations

      Our screened sample had a low number of participants in the older age group. However, it reflects the demographic of this semi-rural regional hub with higher education institutions and professional employment opportunities. We did not specifically collect data on why a participant had attended the hospital precinct for screening. We cannot rule out inclusion of participants who were screened while they were awaiting treatment. The participant medical history was self-reported and not able to be verified against medical records with potential underestimation of the cardiovascular risk factor profile. AF detection rate would be higher with a longer monitoring period. Low health literacy could have contributed to the inability to distinguish incident new AF from prevalent AF, and explained the relatively low proportion of participants on oral anticoagulation therapy. Echocardiogram examination was mostly qualitative due to the nonavailability of fully trained imaging expertise, and we could not confidently ascertain the true proportion of valvular disease due to rheumatic vs degenerative changes. Valvular disease may also have been present in the 6 participants who had AF but did not undertake an echocardiogram.

      Conclusion

      The prevalence of AF from the TEFF-AF screening study in a relatively young semi-rural Ethiopian cohort was <1%. The AF prevalence was higher with increasing age and in those with structural heart disease. Mobile single-lead ECG technology can be used effectively for AF screening in low-resource settings.

      Funding Sources

      Mr Pitman is supported by a Doctoral Scholarship from The Hospital Research Foundation. Dr Wong is supported by a Postdoctoral Fellowship from the National Heart Foundation of Australia. Dr Wong and Dr Lau are supported by Mid-Career Fellowships from The Hospital Research Foundation. Dr Sanders is supported by a Practitioner Fellowship from National Health and Medical Research Council of Australia and by the National Heart Foundation of Australia.

      Disclosures

      Dr Wong reports that the University of Adelaide has received on his behalf lecture, travel, and/or research funding from Abbott Medical, Bayer, Boehringer Ingelheim, Medtronic, Novartis, Servier, and St Jude Medical. Dr Sanders reports having served on the advisory board of Medtronic, Abbott Medical, Boston Scientific, Pacemate, and CathRx. Dr Sanders reports that the University of Adelaide has received on his behalf lecture and/or consulting fees from Medtronic, Abbott Medical, and Boston Scientific. Dr Sanders reports that the University of Adelaide has received on his behalf research funding from Medtronic, Abbott Medical, Boston Scientific, and MicroPort CRM. Dr Lau reports that the University of Adelaide has received on his behalf lecture and/or consulting fees from Abbott Medical, Biotronik, Medtronic, and MicroPort CRM.

      Authorship

      All authors attest they meet the current ICMJE criteria for authorship.

      Patient Consent

      Participants provided informed consent.

      Ethics Statement

      This study received institutional research ethics approval and complied with the ethical principles of the 2013 Declaration of Helsinki.

      Disclaimer

      Given his role as Associate Editor, Dennis Lau had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Editors Jeffrey S. Healey and Jeanne E. Poole.

      References

        • Chugh S.S.
        • Havmoeller R.
        • Narayanan K.
        • et al.
        Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study.
        Circulation. 2014; 129: 837-847
        • Healey J.S.
        • Oldgren J.
        • Ezekowitz M.
        • et al.
        Occurrence of death and stroke in patients in 47 countries 1 year after presenting with atrial fibrillation: a cohort study.
        Lancet. 2016; 388: 1161-1169
        • Stambler B.S.
        • Ngunga L.M.
        Atrial fibrillation in sub-Saharan Africa: epidemiology, unmet needs, and treatment options.
        Int J Gen Med. 2015; 8: 231-242
        • Adedinsewo D.
        • Omole O.
        • Oluleye O.
        • Ajuyah I.
        • Kusumoto F.
        Arrhythmia care in Africa.
        J Interv Card Electrophysiol. 2019; 56: 127-135
        • Harris P.A.
        • Taylor R.
        • Minor B.L.
        • et al.
        The REDCap consortium: building an international community of software platform partners.
        J Biomed Inform. 2019; 95103208
        • Alonso A.
        • Krijthe B.P.
        • Aspelund T.
        • et al.
        Simple risk model predicts incidence of atrial fibrillation in a racially and geographically diverse population: the CHARGE-AF consortium.
        J Am Heart Assoc. 2013; 2e000102
        • Pitman B.M.
        • Chew S.H.
        • Wong C.X.
        • et al.
        Performance of a mobile single-lead electrocardiogram technology for atrial fibrillation screening in a semirural african population: insights from "The Heart of Ethiopia: Focus on Atrial Fibrillation" (TEFF-AF) study.
        JMIR Mhealth Uhealth. 2021; 9e24470
        • Lau J.K.
        • Lowres N.
        • Neubeck L.
        • et al.
        iPhone ECG application for community screening to detect silent atrial fibrillation: a novel technology to prevent stroke.
        Int J Cardiol. 2013; 165: 193-194
        • Freedman B.
        • Camm J.
        • Calkins H.
        • et al.
        Screening for atrial fibrillation: a report of the AF-SCREEN international collaboration.
        Circulation. 2017; 135: 1851-1867
        • Misganaw A.
        • Naghavi M.
        • Walker A.
        • et al.
        Progress in health among regions of Ethiopia, 1990–2019: a subnational country analysis for the Global Burden of Disease Study 2019.
        Lancet. 2022; 399: 1322-1335
        • Tegene E.
        • Tadesse I.
        • Markos Y.
        • Gobena T.
        Prevalence and risk factors for atrial fibrillation and its anticoagulant requirement in adults aged >/=40 in Jimma Town, Southwest Ethiopia: A community based cross-sectional study.
        Int J Cardiol Heart Vasc. 2019; 22: 199-204
        • Giebel G.D.
        • Gissel C.
        Accuracy of mHealth devices for atrial fibrillation screening: systematic review.
        JMIR Mhealth Uhealth. 2019; 7e13641
        • Soni A.
        • Karna S.
        • Patel H.
        • et al.
        Study protocol for Smartphone Monitoring for Atrial fibrillation in Real-Time in India (SMART-India): a community-based screening and referral programme.
        BMJ Open. 2017; 7e017668
        • Halcox J.P.J.
        • Wareham K.
        • Cardew A.
        • et al.
        Assessment of remote heart rhythm sampling using the AliveCor heart monitor to screen for atrial fibrillation: the REHEARSE-AF study.
        Circulation. 2017; 136: 1784-1794
        • William A.D.
        • Kanbour M.
        • Callahan T.
        • et al.
        Assessing the accuracy of an automated atrial fibrillation detection algorithm using smartphone technology: the iREAD study.
        Heart Rhythm. 2018; 15: 1561-1565
        • Turakhia M.P.
        • Desai M.
        • Hedlin H.
        • et al.
        Rationale and design of a large-scale, app-based study to identify cardiac arrhythmias using a smartwatch: the Apple Heart Study.
        Am Heart J. 2019; 207: 66-75
        • Guo Y.
        • Wang H.
        • Zhang H.
        • et al.
        Mobile photoplethysmographic technology to detect atrial fibrillation.
        J Am Coll Cardiol. 2019; 74: 2365-2375
        • Lau D.H.
        • Nattel S.
        • Kalman J.M.
        • Sanders P.
        Modifiable risk factors and atrial fibrillation.
        Circulation. 2017; 136: 583-596
        • Yuyun M.F.
        • Sliwa K.
        • Kengne A.P.
        • Mocumbi A.O.
        • Bukhman G.
        Cardiovascular diseases in sub-Saharan Africa compared with high-income countries: an epidemiological perspective.
        Glob Heart. 2020; (15:15)
        • Linz D.
        • Sanders P.
        • Pitman B.
        • Dobrev D.
        • Lau D.H.
        Atrial fibrillation in sub-Saharan Africa: The knowns and unknowns?.
        Int J Cardiol Heart Vasc. 2019; 22: 212-213
        • Noubiap J.J.
        • Nyaga U.
        • Ndoadoumgue A.
        • Nkeck J.
        • Ngouo A.
        • Bigna J.J.
        Meta-analysis of the incidence, prevalence, and correlates of atrial fibrillation in rheumatic heart disease.
        Global Heart. 2020; 15: 38
        • Yadeta D.
        • Semeredin N.
        • Mekonnen G.
        Prevalence and predictors of atrial fibrillation and its embolic complications in patients with rheumatic heart disease at Tikur Anbessa Specialized Hospital, Addis Ababa, Ethiopia.
        Ethiop J Health Dev. 2019; 33: 12-16
        • Tibazarwa K.
        • Ntyintyane L.
        • Sliwa K.
        • et al.
        A time bomb of cardiovascular risk factors in South Africa: results from the Heart of Soweto Study "Heart Awareness Days".
        Int J Cardiol. 2009; 132: 233-239
        • Virani S.S.
        • Alonso A.
        • Benjamin E.J.
        • et al.
        Heart Disease and Stroke Statistics—2020 Update: a report from the American Heart Association.
        Circulation. 2020; 141: e139-e596
        • Hendriks J.M.
        • Gallagher C.
        • Middeldorp M.E.
        • Lau D.H.
        • Sanders P.
        Risk factor management and atrial fibrillation.
        Europace. 2021; 23 (ii52–ii60)