Wireless Smartphone ECG Enables Large-Scale Screening
Wireless Smartphone ECG Enables Large-Scale Screening
Table 1 summarizes study participant characteristics. Athletes were USC Division I athletes (track + field, rowing, lacrosse, swimming and diving, water polo, volleyball, baseball, cross country, tennis). Healthy young adults were first- and second-year medical students. The patient cohort was comprised of USC ambulatory cardiology patients (age 59 ± 15 years). Heart rates were lowest in athletes and highest in cardiology clinic patients.
Table 2 summarizes the variations in heart rate and conduction interval measurements between the two acquisition methods. Depending on the interval measured, correlation coefficients varied from 0.06 to 0.85. However, absolute differences in conduction intervals were small. Comparing the two acquisition methods, corrected QT interval differed most, followed by PR interval and QRS duration. Heart rate had the least variation. Across the 3 cohorts, absolute mean differences and standard deviations were comparable, with the exception of slightly larger variations in the QRS duration and corrected QT interval in cardiology clinic patients. The most marked differences were seen in paced patients.
Table 3 summarizes the sensitivity and specificity of physician-interpreted smartphone ECGs for the detection of major abnormalities. For the entire population, the smartphone ECG provided accurate detection of atrial rate and rhythm, AV block, and QRS delay; sensitivities ranged from 72.4% (QRS delay) to 94.4% (AF) and specificities were all above 94%.
All AF episodes detected on 12-lead occurred in cardiology clinic patients, and all but one participant was also detected with smartphone ECG. Adjudicating electrophysiologists also detected AF in one athlete and one cardiology clinic patient using the smartphone ECG, but both were found to be in normal sinus rhythm on the 12-lead (false positives). The remainder of the participants in the study had normal sinus rhythm, sinus bradycardia, sinus tachycardia, or junctional rhythm.
The majority of participants with AV block were detected using the smartphone ECG. No participant had advanced AV block. One athlete was found to have Wenckebach conduction, the rest of the participants with block had first degree AV block. Electrophysiologists identified 13 false positive readings for first degree AV block on smartphone ECGs. The majority of AV block was observed in the cardiology clinic patients.
The smartphone ECG allowed electrophysiologists to detect the majority of participants with QRS delay, the bulk of which was observed in cardiology clinic patients. Two healthy young adults were found to have QRS delay on 12-lead but not smartphone ECG (false negatives). False positives were identified in 7 athletes, 2 young adults, and 9 cardiology clinic patients. Figure 1 demonstrates smart phone ECG findings in each of the 3 study cohorts.
(Enlarge Image)
Figure 1.
A: Mobitz type I AV block in an asymptomatic athlete. B: Sinus bradycardia in a medical student. C: Bundle branch block in a cardiology clinic patient. D: Atrial fibrillation in a cardiology clinic patient.
Figure 2 summarizes the postparticipation survey findings. The vast majority of subjects found the wireless ECG more convenient and comfortable. Although only 17% reported having a concern about their heart, the majority felt that using the application would make them feel more secure about their health. Furthermore, most participants wanted to use technology to better communicate with physicians, and the majority predicted they would use the smartphone ECG at least once per month.
(Enlarge Image)
Figure 2.
Athlete and healthy young adult postparticipation survey. app = application.
Results
Participant Demographics
Table 1 summarizes study participant characteristics. Athletes were USC Division I athletes (track + field, rowing, lacrosse, swimming and diving, water polo, volleyball, baseball, cross country, tennis). Healthy young adults were first- and second-year medical students. The patient cohort was comprised of USC ambulatory cardiology patients (age 59 ± 15 years). Heart rates were lowest in athletes and highest in cardiology clinic patients.
Baseline Intervals (Automated Calculations)
Table 2 summarizes the variations in heart rate and conduction interval measurements between the two acquisition methods. Depending on the interval measured, correlation coefficients varied from 0.06 to 0.85. However, absolute differences in conduction intervals were small. Comparing the two acquisition methods, corrected QT interval differed most, followed by PR interval and QRS duration. Heart rate had the least variation. Across the 3 cohorts, absolute mean differences and standard deviations were comparable, with the exception of slightly larger variations in the QRS duration and corrected QT interval in cardiology clinic patients. The most marked differences were seen in paced patients.
Sensitivity and Specificity
Table 3 summarizes the sensitivity and specificity of physician-interpreted smartphone ECGs for the detection of major abnormalities. For the entire population, the smartphone ECG provided accurate detection of atrial rate and rhythm, AV block, and QRS delay; sensitivities ranged from 72.4% (QRS delay) to 94.4% (AF) and specificities were all above 94%.
Atrial fibrillation
All AF episodes detected on 12-lead occurred in cardiology clinic patients, and all but one participant was also detected with smartphone ECG. Adjudicating electrophysiologists also detected AF in one athlete and one cardiology clinic patient using the smartphone ECG, but both were found to be in normal sinus rhythm on the 12-lead (false positives). The remainder of the participants in the study had normal sinus rhythm, sinus bradycardia, sinus tachycardia, or junctional rhythm.
AV Block
The majority of participants with AV block were detected using the smartphone ECG. No participant had advanced AV block. One athlete was found to have Wenckebach conduction, the rest of the participants with block had first degree AV block. Electrophysiologists identified 13 false positive readings for first degree AV block on smartphone ECGs. The majority of AV block was observed in the cardiology clinic patients.
QRS Delay
The smartphone ECG allowed electrophysiologists to detect the majority of participants with QRS delay, the bulk of which was observed in cardiology clinic patients. Two healthy young adults were found to have QRS delay on 12-lead but not smartphone ECG (false negatives). False positives were identified in 7 athletes, 2 young adults, and 9 cardiology clinic patients. Figure 1 demonstrates smart phone ECG findings in each of the 3 study cohorts.
(Enlarge Image)
Figure 1.
A: Mobitz type I AV block in an asymptomatic athlete. B: Sinus bradycardia in a medical student. C: Bundle branch block in a cardiology clinic patient. D: Atrial fibrillation in a cardiology clinic patient.
Athlete and Healthy Young Adult Postparticipation Survey Findings
Figure 2 summarizes the postparticipation survey findings. The vast majority of subjects found the wireless ECG more convenient and comfortable. Although only 17% reported having a concern about their heart, the majority felt that using the application would make them feel more secure about their health. Furthermore, most participants wanted to use technology to better communicate with physicians, and the majority predicted they would use the smartphone ECG at least once per month.
(Enlarge Image)
Figure 2.
Athlete and healthy young adult postparticipation survey. app = application.
Source...