Fig 1.
Schematic view of the Healthdot.
Reprinted from Philips Electronic Nederland BV under a CC BY license, with permission from Philips Electronic Nederland BV, original copyright 2020.
Table 1.
Patient demographics.
Fig 2.
Example of HeartR vitals showing good agreement.
Reference standard (solid line) and Healthdot (dotted line) in bpm.
Fig 3.
Example of HeartR vitals including outliers.
Reference standard (solid line) and Healthdot (dotted line) in bpm.
Fig 4.
Bland-Altman plot of the HeartR.
The difference between the two methods (Healthdot and patient monitor) is plotted against the average of the two, respectively on the y-axis and x-axis. The bias (-0.80 bpm) is indicated by the gray solid line and the confidence interval [CI: 17.8; -19.3 bpm] is indicated by the gray dashed lines.
Fig 5.
Correlation plot of the HeartR.
The reference data (x-axis) is plotted against the Healthdot data (y-axis). The corresponding Pearson correlation coefficient is 0.72 (CI: [0.71:0.72], p < 0.001).
Fig 6.
Error bar of each patient for a 1-sec-interval.
The mean differences and confidence interval (bias +/- 1.96*SD) for each patient are plotted. Difference was calculated by subtracting patient monitor data from Healthdot data, based on a1-sec-interval. The gray dashed lines indicate the required threshold of 5 bpm.
Fig 7.
Error bar of each patient for a 5-min-interval.
The mean differences and confidence interval (bias +/- 1.96*SD) for each patient are plotted. Difference was calculated by subtracting patient monitor data from Healthdot data, based on 5-min averages. The gray dashed lines indicate the required threshold of 5 bpm.
Table 2.
Percentage of patients who met the threshold of 5 bpm for both the mean differences as well as CI for 1-sec averages and 5-min averages.
Fig 8.
Example of RespR vitals showing good agreement.
Reference standard (solid line) and Healthdot (dotted line) in rpm.
Fig 9.
Example of RespR vitals suboptimal agreement.
Reference standard (solid line) and Healthdot (dotted line) in rpm.
Fig 10.
Bland-Altman plot of the RespR.
The difference between the two methods (Healthdot and patient monitor) is plotted against the average of the two, respectively on the y-axis and x-axis. The bias (1.3 bpm) is indicated by the gray solid line and the confidence interval [CI: 8.2; -5.6 bpm] is indicated by the gray dashed lines.
Fig 11.
Correlation plot of the RespR.
The reference data (x-axis) is plotted against the Healthdot data (y-axis). The corresponding Pearson correlation coefficient is 0.64 (CI: of [0.636: 0.644], p < 0.001).
Fig 12.
Error bar of each patient for a 1-sec-interval.
The mean differences and confidence interval (bias +/- 1.96*SD) for each patient are plotted. Difference was calculated by subtracting patient monitor data from Healthdot data, based on a 1-sec-period. The gray dashed lines indicate the required threshold of 5 rpm.
Fig 13.
Error bar of each patient for a 5-min-interval.
The mean differences and confidence interval (bias +/- 1.96*SD) for each patient are plotted. Difference was calculated by subtracting patient monitor data from Healthdot data, based on 5-min averages. The gray dashed lines indicate the required threshold of 5 rpm.
Table 3.
Percentage of patients who met the threshold of 5 rpm for both the mean differences as well as CI for 1-sec averages and 5-min averages.