Hartwacht Arrhythmia (HA) is a cardiac arrhythmia remote monitoring program in the Netherlands, initiated by Cardiologie Centra Nederland. This is the first evaluation of KardiaMobile in a real-world cohort of ambulatory patients for symptom-driven monitoring in that country. Between January 2017 and March 2018, 5,982 KardiaMobile ECGs from 233 participants were received, with a median of 28 ECGs per patient per year (mean age 58 years, 52% male); patients were instructed to record an ECG when they experienced palpitations or related complaints. KardiaMobile algorithms classified 59% as Normal, 22% as Possible AF, 17% as Unclassified, and 2% as Unreadable. According to the HA team, 8% of all ECGs were uninterpretable. The AF algorithm had a sensitivity of 92% and specificity of 95%; the negative predictive value (NPV) for detection of AF was high, at 98%, while the positive predictive value (PPV) was 80%, with 12% of AF ECGs being interpreted by the cardiologist as sinus rhythm. Conversely, the Normal algorithm had a high PPV, and a specificity of 91% and a sensitivity of 80%; 96% of all KardiaMobile Normal ECGs were interpreted by the cardiologist as being sinus rhythm. The authors call for a refinement in the detection of normal sinus rhythm with and without ectopy to reduce the need for manual assessment of this category of ECGs.

Selder JL, Breukel L, Blok S, van Rossum AC, Tulevski II, Allaart CP.Neth Heart J. 2019;27(1):38-45.

This case report describes a woman with congenital pulmonary valve stenosis treated with balloon valvulotomy, with several years of palpitations and unrevealing Holter monitor evaluations. She was enrolled in a remote patient monitoring program in the Netherlands using KardiaPro. The woman used KardiaMobile to successfully record atrial fibrillation with intermittent bundle branch block during an episode of palpitations.

Koole MAC, Somsen GA, Tulevski, II, Winter MM, Bouma BJ, Schuuring MJ.Neth Heart J. Jan 2019.

Non-modifiable risk factors are unchangeable characteristics of a person that make them more vulnerable to developing a cardiovascular disease. Although these risk factors cannot be altered, being aware of them can be helpful in managing your cardiovascular disease risk. Some of the most common non-modifiable risk factors are listed below:ccccc

Koole MAC, Kauw D, Winter MM, Dohmen DAJ, Tulevski, II, de Haan R, et al.Neth Heart J. 2019;27(1):30-7.

KardiaMobile was used in a community screening of 109 patients (70 in sinus rhythm and 39 in AF) soon after a 12-lead ECG had been performed. The ECGs were interpreted by two cardiologists blinded to the rhythm diagnosis, and were processed to provide an automated diagnosis of sinus rhythm or AF. Results were compared with the 12-lead ECG diagnosis by a third cardiologist. An optimized algorithm performed extremely well in the validation set with high sensitivity, specificity, overall accuracy and Kappa (95% CI) of 98% (89%–100%), 97% (93%–99%), 97% (94%–99%) and 0.92 (0.86– 0.98) respectively. This study concluded that KardiaMobile can be used to simply and rapidly record a high quality single-lead ECG to accurately detect AF, making it an ideal technology for community screening programs to detect silent AF.

Lau JK, Lowres N, Neubeck L, Brieger DB, Sy RW, Galloway CD, et al.Int J Cardiol. 2013;165(1):193-4.

One thousand pharmacy customers (mean age 76 ± 7 years, 44% male) were screened with KardiaMobile. Newly identified AF was found in 1.5% (95% CI, 0.8-2.5%), and AF prevalence was 6.7%. The automated algorithm showed 98.5% sensitivity and 91.4% specificity for detecting AF. Using cost and outcome data from a United Kingdom study for AF screening, the incremental cost-effectiveness ratio of extending screening into the community with KardiaMobile, based on 55% warfarin prescription adherence, would be $USD4,066 per quality-adjusted life-year gained, and $USD20,695 for preventing one stroke. In summary, screening for AF with KardiaMobile is feasible and cost-effective.

Lowres N, Neubeck L, Salkeld G, Krass I, McLachlan AJ, Redfern J, et al.Thromb Haemost. 2014;111(6):1167-76.

KardiaMobile was used to identify asymptomatic AF at the time of influenza vaccination in 5 practices in Sydney, Australia. Nurses used the automated algorithm to screen 973 patients aged ≥ 65 years between April-June 2015. Screening took on average 5 minutes (range 1.5 -10 minutes); abnormal recordings required additional time. Newly identified AF was found in 0.8% (8) of patients, and the overall prevalence of AF was 3.8% (37). The sensitivity and specificity of the automated algorithm for detecting AF was 95% and 99%, respectively. Screening by practice nurses was well accepted by practice staff. Key enablers were the confidence and competence of nurses and a ‘designated champion’ to lead screening at the practice. Barriers were practice specific, and mainly related to staff time and funding.

Orchard J, Lowres N, Freedman SB, Ladak L, Lee W, Zwar N et al.Eur J Prev Cardiol. 2016; 23(2S): 13-20.

Heart rate (HR) detection from a smartphone-based photoplethysmography (PPG) app (FibriCheck) was compared with the KardiaMobile ECG and the Nonin pulse oximeter. The HR (BPM, beats per minute) of 88 random subjects consecutively measured for 10 seconds with the 3 devices showed a moderate-to-strong correlation coefficient of 0.834 between FibriCheck and Nonin, 0.88 between FibriCheck and AliveCor, and 0.897 between Nonin and AliveCor. The mean HR for FibriCheck was 71 BPM, for Nonin 69 BPM, and for AliveCor 69 BPM. A single way analysis of variance showed no significant differences between the HRs as measured by the 3 devices (p=0.61). This study reports the potential utility and limitations in use of the smartphone-based PPG signal for HR detection.

Vandenberk T, Stans J, Van Schelvergem G, Pelckmans C, Smeets CJ, Lanssens D, et al.JMIR Mhealth Uhealth. 2017;5(8):e129.

The accuracy of the KardiaMobile AF algorithm was evaluated in 52 patients admitted for antiarrhythmic drug initiation for AF. Patients performed KardiaMobile recordings immediately following twice daily 12-lead ECGs. There were 225 paired KardiaMobile and 12-lead ECG recordings. The algorithm interpretation was missing or labeled as non-interpretable in 62 (27.5%) of recordings for multiple reasons (truncated recording, noise, slow heart rate, other). When the algorithm did not provide a diagnosis, blinded electrophysiologists were able to provide interpretation in 92% of these recordings. After exclusion of non-interpretable recordings, the KardiaMobile AF algorithm had very good accuracy, with a sensitivity of 96.6% and specificity of 94% for the detection of AF when compared to physician interpreted ECGs, and a κ coefficient of 0.89. The majority of patients (93.6%) found KardiaMobile easy to use, and 59.6% noted that use lessened AF-diagnosis related anxiety. 63.8% of survey respondents preferred continued use of KardiaMobile for AF detection.

William AD, Kanbour M, Callahan T, Bhargava M, Varma N, Rickard J, et al.Heart Rhythm. August 2018. doi: 10.1016/j.hrthm.2018.06.037

The accuracy of the KardiaMobile AF algorithm was evaluated in 52 patients admitted for antiarrhythmic drug initiation for AF. Patients performed KardiaMobile recordings immediately following twice daily 12-lead ECGs. There were 225 paired KardiaMobile and 12-lead ECG recordings. The algorithm interpretation was missing or labeled as non-interpretable in 62 (27.5%) of recordings for multiple reasons (truncated recording, noise, slow heart rate, other). When the algorithm did not provide a diagnosis, blinded electrophysiologists were able to provide interpretation in 92% of these recordings. After exclusion of non-interpretable recordings, the KardiaMobile AF algorithm had very good accuracy, with a sensitivity of 96.6% and specificity of 94% for the detection of AF when compared to physician interpreted ECGs, and a κ coefficient of 0.89. The majority of patients (93.6%) found KardiaMobile easy to use, and 59.6% noted that use lessened AF-diagnosis related anxiety. 63.8% of survey respondents preferred continued use of KardiaMobile for AF detection.

William AD, Kanbour M, Callahan T, Bhargava M, Varma N, Rickard J, et al.Heart Rhythm. August 2018. doi: 10.1016/j.hrthm.2018.06.037

A paucity of data exists on the accuracy of primary care physicians’ (PCP) interpretation of KardiaMobile ECGs compared with the device’s automated diagnosis. Using 408 ECGs in 51 patients, before and after elective cardioversion, this study demonstrated variable accuracy in clinician interpretation, with a mean accuracy of 91% for the review by cardiologists, and 85% accuracy for the review by PCPs. With exclusion of Unclassified ECGs, the algorithm accuracy had a sensitivity and specificity of 100% and 95%, respectively. Accurate diagnosis of a KardiaMobile Unclassified ECG was established in 10/12 when assessed by a cardiologist, and 9/12 on review by a primary care physician. Combining the automated algorithm with cardiologist interpretation of only Unclassified traces yielded excellent results and provides an efficient, cost-effective workflow for the utilization of a smartphone-based ECG in clinical practice.

Koshy AN, Sajeev JK, Negishi K, Wong MC, Pham CB, Cooray SP, et al.Am Heart J. 2018. doi: 10.1016/j.ahj.2018.08.001.

This study evaluated the accuracy of KardiaBand ECG and the automated AF algorithm. 100 patients (mean age 68 ± 11 yrs) with AF presenting for cardioversion (CV) were enrolled and received simultaneous 12-lead ECG and KardiaBand ECG before the procedure; if the CV was performed a post CV 12-lead ECG was then obtained along with another KardiaBand ECG. CV was canceled in 8 patients due to presentation in sinus rhythm. There were 169 simultaneous 12-lead ECG and KardiaBand ECGs. Compared to 12-lead ECG, the automated algorithm detected AF with 93% sensitivity, 84% specificity and K coefficient 0.77. Physician-interpretation of KardiaBand ECGs demonstrated 99% sensitivity, 83% specificity and K coefficient 0.83. The automated AF algorithm on KardiaBand, when supported by physician review, can accurately differentiate AF from sinus rhythm. This technology can help screen patients prior to elective CV and avoid unnecessary procedures.

Bumgarner JM, Lambert CT, Hussein AA, Cantillon DJ, Baranowski B, Wolski K, et al.JACC. March 2018. DOI:10.1016/j.jacc.2018.03.003

This study evaluated the accuracy of KardiaBand ECG and the automated AF algorithm. 100 patients (mean age 68 ± 11 yrs) with AF presenting for cardioversion (CV) were enrolled and received simultaneous 12-lead ECG and KardiaBand ECG before the procedure; if the CV was performed a post CV 12-lead ECG was then obtained along with another KardiaBand ECG. CV was canceled in 8 patients due to presentation in sinus rhythm. There were 169 simultaneous 12-lead ECG and KardiaBand ECGs. Compared to 12-lead ECG, the automated algorithm detected AF with 93% sensitivity, 84% specificity and K coefficient 0.77. Physician-interpretation of KardiaBand ECGs demonstrated 99% sensitivity, 83% specificity and K coefficient 0.83. The automated AF algorithm on KardiaBand, when supported by physician review, can accurately differentiate AF from sinus rhythm. This technology can help screen patients prior to elective CV and avoid unnecessary procedures.

Bumgarner JM, Lambert CT, Hussein AA, Cantillon DJ, Baranowski B, Wolski K, et al.JACC. March 2018. DOI:10.1016/j.jacc.2018.03.003

148 patients (mean age 41 years) with intermittent palpitations were asked to use KardiaMobile and record an ECG when symptomatic. Over a median period of use of 244 days, 113 patients (76.4%) made 516 symptomatic recordings. A symptom-rhythm correlation was possible for all patients who submitted recordings. Diagnoses were: sinus rhythm n=47 (41.6%), sinus tachycardia n=21 (18.6%), supraventricular/ventricular ectopics n=31 (27.4%), atrial fibrillation n=8 (7.1%), and supraventricular tachycardia n=6 (5.3%). Median time to diagnosis was nine days (range 1–287 days). In conclusion, KardiaMobile diagnosed the cause of intermittent palpitations in the majority of patients referred for evaluation.

Dimarco AD, Onwordi EN, Murphy CF, Walters EJ, Willis L, Mullan NJ, et al.Brit J Cardio. March 2018. doi:10.5837/bjc.2018.006

This study evaluated the between and within rater validity and reliability of KardiaMobile in recording ECG rate, rhythm, and intervals in healthy college athletes in a pre-participation screening program. First, 10 athletes’ KardiaMobile ECG were reviewed by three physicians. Second, the physicians compared a 30-second simultaneous KardiaMobile and lead I from a 12-lead ECG, from 12 athletes. The between rater and between device reliability for the rate, QT interval and QRS duration parameters ranged from good to very good (intraclass correlation coefficients [ICC] = 0.667 – 0.981). The current investigation showed that the reliability of the ECG parameters measured using the smartphone technology ranged from good to very good. This paper serves as support for a technological advancement that will help advance the debate on the utility of ECG testing as part of the athletic pre-participation physical.

Gilliland A, Timmons M, Harris JK, Petrany SM, Shepherd GS, Buchanan GS, et al.Marshall J Med. 2018;4(2):61-74.

This study evaluated the use KardiaMobile in the lead II position (right hand to left leg) to improve visualization of flutter waves and clinician diagnosis of atrial flutter (AFL), compared to traditional lead-I tracings. Fifty patients were recruited (25 in sinus rhythm, 14 in AF, 11 in AFL). Lead-I AFL sensitivity was 27% for both electrophysiologists (EP), which individually improved to 73% and 55% in lead-II. KardiaMobile appropriately diagnosed lead-I AFL as unclassified in 18% of cases, compared to 55% in lead-II. Overall clinician agreement (AF, sinus rhythm and AFL) was modest utilizing lead-I position (EP1: κ=0.71, EP2: κ=0.73, p<0.001), which improved with lead-II tracings (EP1: κ=0.87, EP2: κ=0.83, both p<0.001). In summary, lead II position of KardiaMobile improves clinician diagnosis of atrial flutter.

Rajakariar K, Koshy AN, Sajeev JK, Nair S, Roberts L, Teh AW.J Electrocardiol. 2018. 51(5): 884-88.

This is the first study to evaluate the use of KardiaMobile in the urgent care setting. Those seeking care at urgent care centers constitute the fastest growing segment of patients in the U.S., and appropriately triaging those with palpitations will positively affect the entire American healthcare system. The purpose of this study is to compare KardiaMobile for 30 days with 24-hour Holter monitoring for the detection of symptomatic arrhythmias in an urgent care population. All Holter reports and KardiaMobile ECGs were reviewed by a general practitioner and a cardiologist. Data from the first 50 of 100 patients from 6 urgent care centers across Southern Arizona was performed. KardiaMobile was diagnostically superior to (10%) or concordant with (72%) Holter monitoring in 82% of patients. Holter monitoring was superior in 16% of patients. Arrhythmias detected included atrial and ventricular ectopy, supraventricular tachycardia, atrial fibrillation and inappropriate sinus tachycardia. This ongoing study will ultimately analyze noninferiority for diagnosis and management, perform a cost-comparison, and suggest the most effective clinical uses for KardiaMobile in an urgent care setting.

Goel HV, Alpert JS, Shaheen MH, Jones TA, and Skinner DP.American Heart Association Scientific Sessions (2018). Abstract.

In this Viewpoint, the author describes the potential utility of photoplethymographic (PPG) sensors in consumer-grade wearable devices, such as Fitbit or Apple Watch, to aid in detection of arrhythmias. He also opines that it is essential to validate abnormal PPG findings with direct electrocardiographic recordings, such as from KardiaMobile or Apple Watch series 4. The opinion piece includes a figure showing the use of KardiaMobile ECG in the lead I and lead II orientations during normal sinus rhythm and during supraventricular tachycardia (long RP tachycardia).

Ip JE.JAMA. Online Jan 2019.

This case report presents a patient enrolled in a precision medicine study in which remote patient monitoring helped detect the presence of atrial fibrillation-atrial flutter(AFib-Flutter). A 64-year-old male with a history of ischemic heart disease used KardiaMobile after noting chest pain, and received an instant analysis of Atrial Fibrillation with a heart rate of 139 bpm. He called the paramedics and was taken by ambulance to the emergency department, where AFib-Flutter with rapid ventricular response was confirmed by 12-lead ECG, and successfully treated.

Joung S, Dzubur E, van den Broek I, Love A, Martinez-Rubio L, Lopez M, et al.J Med Cases. 2019; 10(2):31-36.

Palpitations and pre-syncope are together responsible for 300,000 annual Emergency Department (ED) attendances in the United Kingdom (UK) alone. This multicenter randomized controlled trial compared the symptomatic rhythm detection rate of KardiaMobile versus standard care alone (no planned ambulatory ECG monitoring), for 243 participants presenting to 10 emergency departments in the UK with palpitations and pre-syncope with no obvious cause evident at initial consultation. A symptomatic rhythm was detected at 90 days in 69 (n=124; 55.6%; 95% CI 46.9–64.4%) participants in the intervention group versus 11 (n=116; 9.5%; 95% CI 4.2–14.8) in the control group (RR 5.9, 95% CI 3.3–10.5; p<0.0001). Mean time to symptomatic rhythm detection in the intervention group was 9.5 days (SD 16.1, range 0–83) versus 42.9 days (SD 16.0, range 12–66; p<0.0001) in the control group. Use of KardiaMobile increased the number of patients with symptomatic rhythm detection over five-fold, to more than 55%, at 90 days. The authors recommend that KardiaMobile be considered part of on-going care to all patients presenting acutely with unexplained palpitations or pre-syncope.

Reed MJ, Grub NR, Lang CC, O’Briend R, Simpson K, Padarenga M, et al.EClinicalMedicine. Online March 3, 2019.

This guidance document for the United Kingdom evaluated use of lead-I ECG devices for single time point testing of people in primary care with symptoms of atrial fibrillation and an irregular pulse. The authors concluded there is not enough evidence to recommend routine adoption of lead-I ECG devices for this use case. They recommended further research to show how using lead-I ECG affects the number of people with atrial fibrillation detected, as well the staff time needed to interpret the ECG tracings. Of note, a de novo economic model was designed to evaluate cost effectiveness, and KardiaMobile dominated all other lead-I ECG devices, costing less and producing more quality-adjusted life years [QALYs].

National Institute for Health and Care Excellence. May 2019.

Diagnostic pathways for identification of clinically significant paroxysmal arrhythmia have historically relied on ambulatory ECG monitoring. While useful as a risk stratification tool in certain patient groups, in general, it has a limited yield for infrequent arrhythmia. This is inherently cost-inefficient, and time to diagnosis can be delayed. Smartphone-based ECG devices are now well established in the public market. However, their adoption into standard investigatory pathways is not yet widespread. Recently, the National Institute for Health and Care Excellence (NICE) published a diagnostics guidance document, which reviewed the smartphone-based devices, KardiaMobile and imPulse, with respect to atrial fibrillation (AF) detection. NICE concluded there was insufficient evidence to recommend routine adoption in primary care and recommended further research. Regardless, given the evidence that is available and their uptake by the general public, cardiologists are increasingly likely to encounter them. This article focuses on KardiaMobile as the device is currently easily available to the general public. The authors conclude there is growing evidence to support its use as a screening tool for the detection of subclinical AF in high-risk populations, but further studies are required in order to equate this benefit to stroke and mortality reduction.

Bennett R and French A.Heart. 2019.

In 10 Dutch general practices, KardiaMobile ECG and AF algorithm were compared with simultaneous 12-lead ECG. Three cardiologists reviewed ECG data from 214 patients (mean age 64.1 y, 54% male). The 12-lead ECG diagnosed AF/AFL, any rhythm abnormality, and any conduction abnormality (AV block, BBB, LAD, LAFB) in 23, 44, and 28 patients, respectively. KardiaMobile ECG as assessed by the cardiologists had a sensitivity and specificity for AF/AFL of 100% (95% CI, 85.2%-100%) and 100% (95% CI, 98.1%-100%). The AF Instant Analysis algorithm identified 20 or 23 AF cases and incorrectly classified 4 cases of sinus rhythm as possible AF (sensitivity and specificity of 87.0% (95% CI, 66.4%-97.2%) and 97.9% (95% CI, 94.7%-99.4%)). KardiaMobile recordings as assessed by cardiologists had a sensitivity and specificity for any rhythm abnormality of 90.9% (95% CI, 78.3%-97.5%) and 93.5% (95% CI, 88.7%-96.7%) and for any conduction abnormality of 46.4% (95% CI, 27.5%-66.1%) and 100% (95% CI, 98.0%-100%). For conduction abnormality, the 15 false negatives were comprised of first-degree AVB (n=6), LAFB (n=8), and RBBB (n=1); on the other hand, cardiologists were able to accurately identify BBB in 13 patients’ KardiaMobile ECGs. The authors concluded that in a primary care population, the KardiaMobile ECG recording showed excellent diagnostic accuracy for AF/AFL and good diagnostic accuracy for other rhythm abnormalities. The 1L-ECG device was less sensitive for left anterior fascicular block and first-degree AV block.

Himmelreich JC, Karregat EPM, Lucassen, WAM, van Weert HCPM, de Groot JR, Handoko ML, et al.Ann Fam Med. 2019;17: 403-11.