OBJECTIVES:

Continuous pulse oximetry monitoring is routine in many pediatric inpatient units, generating hundreds of alarms per patient per day, up to 90% of which are false (nonactionable). We hypothesized that continuously monitoring pulse oximetry with a partially wireless monitor without physical connection to a wall unit (wireless pox) will decrease episodes of loss of signal integrity (LOSI) generated by motion artifact in healthy pediatric subjects completing age-appropriate activities compared with a traditional wall-connected pulse oximeter (control pox), thereby reducing false alarms.

METHODS:

Thirty-six healthy children, aged 1 to 17 years, were continuously monitored simultaneously with a wireless pox and a control pox while performing increasing levels of activity for 20 minutes. Continuous pulse oximetry data were recorded in 2 second intervals. Data were analyzed for LOSI. The Wilcoxon signed rank test was then used to compare the control pox to the wireless pox.

RESULTS:

The wireless pox had fewer mean number of alarms due to LOSI (control 7.86, wireless 4.17, P = .0031) and fewer mean episodes of LOSI not leading to alarms (control 9.94, wireless 6.92, P = .0006). The control pox had a longer percent time in alarm state related to LOSI in all age groups.

CONCLUSIONS:

This prospective observational study found that the wireless pox has decreased number and duration of events of LOSI compared with the control pox in healthy children. Implementation of partially wireless pulse oximetry in pediatric inpatient units may decrease episodes of LOSI and false alarms from motion artifact.

Continuous pulse oximetry monitoring remains routine in many inpatient pediatric units and is a source of frequent hospital alarms, generating 39 to 352 alarms per patient per day, up to 90% of which are false (nonactionable).1  Alarm fatigue has become a top patient safety concern. As of January 1, 2016, the Joint Commission required all hospitals to establish policies and procedures for managing alarm fatigue.2  Many novel ways to address alarm fatigue have been proposed including changing alarm parameters on the basis of specific patient characteristics and adding adaptive time delays.3,4  Yet the number of actionable alarms continues to be low, estimated to be between <1% and 26% of all hospital alarms.5  In 1 study, researchers demonstrated that 44% of all alarms in a PICU originated from pulse oximeters. The same study showed that 71% of the pulse oximeter alarms were false and only 7% of the pulse oximeter alarms were clinically significant.6  Reduction of invalid alarms is 1 method of reducing alarm fatigue.1 

One well-known etiology of erroneous pulse oximetry readings is motion artifact.79  In addition to causing false alarms, motion artifact can lead to misleading device output without triggering an alarm state. A brief review of how pulse oximetry works is helpful in understanding the effects of motion on pulse oximetry. Pulse oximetry depends on detection of light passed through tissue, often a finger or toe. To work properly, the emitter must be directly lined up with the sensor with the tissue being measured sandwiched between them. The detected light is then analyzed by using proprietary algorithms within the device and the patient’s pulse oxygen saturation (Spo2) is reported. Pulse oximetry algorithms assume that all pulsatile flow is due to arterial blood. Changes in the detection of pulsatile flow impact the calculation of the oxygen saturation. During movement, venous blood volume may change and be seen as a pulsatile component that is erroneously incorporated into the measurement of the Spo2.7,10  A second cause of loss of signal integrity (LOSI) is from the sensor moving in relation to the patient causing misalignment of the emitter and detector across the tissue bed and erroneous Spo2 measurements. The final form of motion artifact of interest for this study includes episodes of LOSI from signal transmission disruption such as physical interruption of the signal pathway that may be exacerbated by physical activity (for example, the pulse oximeter probe falling off the patient). It is common practice in pediatric inpatient units to continuously monitor pulse oximetry by using devices that connect the sensor probe on the patient to a wall unit using an 8 to 12-ft cable. This long cable may exacerbate motion of the sensor in relation to the patient, which may contribute to episodes of LOSI.

To evaluate the concept that cable length may contribute to LOSI in hospitalized children requiring continuous pulse oximetry monitoring, we began with this pilot study using healthy test subjects to fully optimize motion during testing and to eliminate patient factors that may cause true oxygen desaturation. This prospective study compares LOSI between a traditional wall-connected pulse oximeter (control pox) and a partially wireless monitor without physical connection to a wall unit (wireless pox) in healthy pediatric subjects completing fine and gross motor activities. We hypothesized that continuously monitoring pulse oximetry with a wireless pox would decrease the frequency of episodes of LOSI generated by motion artifact in healthy pediatric subjects completing age-appropriate activities.

We conducted a prospective, observational study on healthy pediatric subjects between the ages of 1 month and 17 years of age. The hospital institutional review board approved this study and informed consent from a legal guardian was obtained for all subjects, with additional assent for subjects aged 7 years or older. Children were recruited by using fliers posted in several inpatient and outpatient areas of the hospital campus. Healthy subjects were enrolled and tested between July 2017 and November 2017. Exclusion criteria included the presence of cardiac or respiratory conditions including asthma with use of respiratory medications in the past 2 weeks, or any use of medications for cardiac conditions. These disease states were excluded because they may be associated with altered Spo2 during rest or activity independent of device-related effects. The study was designed to include the following age groups: 1 to 11 months, 1 to 5, 6 to 10, and 11 to 17 years. These age groups represent the most common developmental groups seen in inpatient pediatric units and were stratified to identify the age groups most likely to benefit from future study. The age group of 1 to 11 months was eventually eliminated secondary to slow enrollment.

Testing occurred in the Translational Research Unit of the hospital designed to mimic outpatient clinic rooms. Each subject was continuously monitored simultaneously with both the partially wireless Radius 7 (Masimo, Irvine, CA) (wireless pox) and the wall-connected Radical 7 (Masimo) (control pox) pulse oximeters. The Radius 7 and Radical 7 and their corresponding soft wrap probes are US Food and Drug Administration approved for continuous monitoring in pediatric patients within the age ranges studied. The internal settings of the wireless pox and control pox were configured in the exact same way including alarm limits, sampling time, and averaging time. The wireless pox was set to use exclusively Bluetooth connectivity to communicate with the ROOT Monitor (Masimo) for display and data collection. The wireless pox monitor and the control pox were located next to each other against the back wall of the room. The control pox was connected to the patient via a 10-ft-long cable, and the device was kept in a stationary location with relation to the child to simulate a hospital room (where the device is usually mounted near the side of a bed). The children were allowed to move freely around the room, so distance from the wireless pox monitor and the control pox varied throughout testing. The wireless pox was attached to either the right or left middle finger preferably, or hand or foot if necessary, for patient tolerance or to accommodate patient size. The control pox was attached to the contralateral middle finger or hand or foot, matching the location of the wireless pox. The side to which the wireless pox was attached first was randomized within each age group on the basis of dominant versus nondominant hand. Halfway through the testing, the devices were switched to the opposite sides to account for any differences in movements secondary to hand dominance or any difference in baseline perfusion of the testing site.

All subjects performed age-appropriate activities as developed by local physical therapists (Supplemental Table 3) for ∼20 minutes total while connected to the pulse oximeters. A baseline Spo2 reading was obtained for both devices during the initial 3 minutes with patient at rest. For the remainder of the study, subjects were encouraged to perform activities at increasing movement levels progressing from occasional fine movements to consistent large-gross movements in 3-minute increments. Halfway through the testing, the probe locations were switched to the contralateral limbs and monitoring continued as subjects performed movement activities as described above. For ease of description, an Activity Likert scale was created to grade the activity levels. On the scale, level 1 was minimal movement; level 2 was consistent fine movements; level 3 was occasional large-gross movements; and level 4 was consistent large-gross movements (Supplemental Table 4). Data were recorded for baseline measurement and 3 different activity scale levels for each location of the monitors. If at any time during testing the pox probe fell off the child, the technician replaced the probe immediately and the child finished the testing.

The continuous pulse oximetry data were downloaded from the control pox and wireless pox devices using the Masimo Instrument Configuration Tool v1060 software (Masimo, Irvine, CA). This provided data every 2 seconds, including date- and time-stamped recordings of Spo2 and all warning and alarm states including reasons for alarms. LOSI events were classified as alarms due to LOSI (complete inability to pick up the patient signal or a change in the pulse oximetry saturation value by 4% for more than 2 consecutive seconds). The change in saturation value of 4% was based on maximal desaturation noted in healthy adults during exercise11  and the 2-second time duration was based on sampling frequency of the devices. Alarms correlating with inability to pick up the patient signal for the purposes of this study were “sensor off patient,” “sensor not connected,” and “Spo2 low” (set in the device as Spo2 <92%). Ninety-two percent was set as a pragmatic cutoff as a surrogate value to detect LOSI because there is no clear established normal Spo2 value for healthy children in the literature, with the definition for hypoxemia ranging from 90% to 97% depending on the study.12,13  LOSI events were also classified as LOSI not leading to alarm state.

Data from each subject were analyzed for the number, duration, and time percentage in each episode of LOSI and for all alarms secondary to LOSI. The number of episodes is the count of uninterrupted, discrete LOSI events, including those leading to alarms. The duration is the time in seconds of an individual LOSI event or alarm. The time percentage is the summed duration of all events divided by the total time of study. Data from individual subjects were then combined, and the Wilcoxon signed rank test was used to compare the wireless pox versus control pox. A P value <.05 was considered statistically significant. Statistical software employed was SAS version 9.4 (SAS Institute, Inc, Cary, NC).

Thirty-six subjects were included in the analysis: 12 in each age group (1–5, 6–10, and 11–17 years). Subjects were 50% male, and 86% were white, 8% were African American, and 6% were Asian American. Percent of testing time for each subject and intensity of activity were not significantly different between the 2 devices.

Overall, the wireless pox had fewer alarms due to LOSI and fewer number of episodes of LOSI not leading to alarms compared with the control pox (Fig 1). The control pox had significantly higher number, longer duration, and more percent time in alarm state due to LOSI for “sensor not connected” compared with the wireless pox (Table 1). The control pox also had longer duration, and more percent time in alarm state due to “sensor off patient.” The warnings for “sensor not connected” and “sensor off patient” clinically corresponded to the sensor not being properly connected to the subject, which occurred when the probe fell off the subject during activity.

FIGURE 1

Mean number of episodes of LOSI with and without alarm.

FIGURE 1

Mean number of episodes of LOSI with and without alarm.

Close modal
TABLE 1

Analysis of Episodes of LOSI

LOSI EventWireless Pox, Median (IQR)Control Pox, Median (IQR)P
“Sensor not connected” alarm, N 0 (0–1) 2 (1–3) <.001 
“Sensor not connected” alarm duration, s 0 (0–2) 39 (20–129) <.001 
“Sensor not connected” % time in alarm 0 (0–0.1) 3 (2–10) <.001 
“Sensor off patient” alarm, N 0 (0–1) 0 (0–2) .12 
“Sensor off patient” alarm duration, s 0 (0–8) 9 (0–39) .036 
“Sensor off patient” % time in alarm 0 (0–0.6) 0.6 (0–3) .028 
“Spo2 decrease >4%” episode, Na 4 (1–10) 8 (4–12) .005 
“Spo2 decrease >4%” episode duration, sa 14 (3–39) 32 (17–59) <.001 
“Spo2 decrease >4%” % time in episodea 1 (0–3) 2 (1–4) <.001 
“Missing values in Spo2” episode, Na 0 (0–2) 2 (1–3) .15 
“Missing values in Spo2” episode duration, sa 0 (0–38) 47 (6–81) .011 
“Missing values in Spo2” % time in episodea 0 (0–3) 3 (0.4–6) .011 
“Spo2 low” alarm, N 2 (0–4) 3 (1–6) .14 
“Spo2 low” alarm duration, s 21 (0–43) 26 (3–107) .13 
“Spo2 low” % time in alarm 1 (0–3) 2 (0.2–7) .12 
LOSI EventWireless Pox, Median (IQR)Control Pox, Median (IQR)P
“Sensor not connected” alarm, N 0 (0–1) 2 (1–3) <.001 
“Sensor not connected” alarm duration, s 0 (0–2) 39 (20–129) <.001 
“Sensor not connected” % time in alarm 0 (0–0.1) 3 (2–10) <.001 
“Sensor off patient” alarm, N 0 (0–1) 0 (0–2) .12 
“Sensor off patient” alarm duration, s 0 (0–8) 9 (0–39) .036 
“Sensor off patient” % time in alarm 0 (0–0.6) 0.6 (0–3) .028 
“Spo2 decrease >4%” episode, Na 4 (1–10) 8 (4–12) .005 
“Spo2 decrease >4%” episode duration, sa 14 (3–39) 32 (17–59) <.001 
“Spo2 decrease >4%” % time in episodea 1 (0–3) 2 (1–4) <.001 
“Missing values in Spo2” episode, Na 0 (0–2) 2 (1–3) .15 
“Missing values in Spo2” episode duration, sa 0 (0–38) 47 (6–81) .011 
“Missing values in Spo2” % time in episodea 0 (0–3) 3 (0.4–6) .011 
“Spo2 low” alarm, N 2 (0–4) 3 (1–6) .14 
“Spo2 low” alarm duration, s 21 (0–43) 26 (3–107) .13 
“Spo2 low” % time in alarm 1 (0–3) 2 (0.2–7) .12 

A P value <.05 is considered statistically significant. IQR, interquartile range.

a

LOSI event without alarm state.

When analyzing both LOSI leading to alarms and LOSI not leading to alarms, wireless pox was found to have better signal integrity overall compared with the control pox (Table 1). The control pox had significantly higher number, longer duration, and more percent time in episodes of LOSI corresponding to “Spo2 decrease >4%” compared with the wireless pox. The control pox also had significantly longer duration and more percent time in episodes of LOSI corresponding to “missing values in Spo2” that did not lead to alarm state compared with the wireless pox.

As shown in Table 2, subgroup analysis by activity level for LOSI leading to alarm and LOSI not leading to alarm found that the control pox had higher number, longer duration, and more percent time spent in LOSI. When analyzing by age group, control pox had more alarms due to LOSI (Fig 2) compared with the wireless pox in the 11 to 17 year age group, and LOSI without alarm (Fig 3) in the 6 to 10 and 11 to 17 year age groups. The control pox had a longer percent of time in alarm state caused by LOSI compared with the wireless pox in all age groups.

TABLE 2

Analysis of Alarms and LOSI Events Not Leading to Alarm State by Activity Level

Alarms for “Sensor Not Connected”Alarms for “Spo2 Low”Events of “Spo2 Decrease >4%”Events of “Missing Values in Spo2
Activity LevelWireless Pox, Median (IQR)Control Pox, Median (IQR)PWireless Pox, Median (IQR)Control Pox, Median (IQR)PWireless Pox, Median (IQR)Control Pox, Median (IQR)PWireless Pox, Median (IQR)Control Pox, Median (IQR)P
No. alarms 0 (0–0) 0 (0–1) .001 0 (0–0) 0 (0–1) .001 0 (0–0) 0 (0–0) .11 0 (0–0) 0 (0–1) .018 
 0 (0–0) 0.5 (0–1) .010 0 (0–0) 0 (0–2) .009 0 (0–1) 0 (0–1) .29 0 (0–0) 0 (0–1) .27 
 0 (0–0) 0 (0–1) <.001 1 (0–2) 0 (0–2) .8 1.5 (0–4) 3 (1–5) .34 0 (0–0) 0 (0–0) .6 
 0 (0–0) 0 (0–1) .018 0 (0–1) 0 (0–2) .87 1 (0–4) 3 (1–7) <.001 0 (0–1) 0 (0–1) .95 
Duration of alarm, s 0 (0–0) 0 (0–24) <.001 0 (0–0) 0 (0–4) <.001 0 (0–0) 0 (0–0) .047 0 (0–0) 0 (0–26) <.001 
 0 (0–0) 1 (0–34) <.001 0 (0–0) 0 (0–19) .010 0 (0–2) 0 (0–2) .36 0 (0–0) 0 (0–23) .047 
 0 (0–0) 0 (0–14) <.001 2 (0–23) 0 (0–17) .65 8 (0–22) 10 (2–20) .49 0 (0–0) 0 (0–0) .86 
 0 (0–0) 0 (0–16) .009 0 (0–13) 0 (0–15) .96 2 (0–12) 17 (2–31) <.001 0 (0–10) 0 (0–19) .24 
Time in alarm, % 0 (0–0) 0 (0–11) <.001 0 (0–0) 0 (0–2) <.001 0 (0–0) 0 (0–0) .094 0 (0–0) 0 (0–13) <.001 
 0 (0–0) 0.3 (0–8) <.001 0 (0–0) 0 (0–5) .009 0 (0–0.6) 0 (0–0.5) .31 0 (0–0) 0 (0–5) .043 
 0 (0–0) 0 (0–4) <.001 0.56 (0–5) 0 (0–4) .96 2 (0–6) 3 (0.5–6) .37 0 (0–0) 0 (0–0) >.99 
 0 (0–0) 0 (0–4) .008 0 (0–4) 0 (0–5) .89 0.6 (0–3) 5 (0.6–8) <.001 0 (0–2) 0 (0–5) .23 
Alarms for “Sensor Not Connected”Alarms for “Spo2 Low”Events of “Spo2 Decrease >4%”Events of “Missing Values in Spo2
Activity LevelWireless Pox, Median (IQR)Control Pox, Median (IQR)PWireless Pox, Median (IQR)Control Pox, Median (IQR)PWireless Pox, Median (IQR)Control Pox, Median (IQR)PWireless Pox, Median (IQR)Control Pox, Median (IQR)P
No. alarms 0 (0–0) 0 (0–1) .001 0 (0–0) 0 (0–1) .001 0 (0–0) 0 (0–0) .11 0 (0–0) 0 (0–1) .018 
 0 (0–0) 0.5 (0–1) .010 0 (0–0) 0 (0–2) .009 0 (0–1) 0 (0–1) .29 0 (0–0) 0 (0–1) .27 
 0 (0–0) 0 (0–1) <.001 1 (0–2) 0 (0–2) .8 1.5 (0–4) 3 (1–5) .34 0 (0–0) 0 (0–0) .6 
 0 (0–0) 0 (0–1) .018 0 (0–1) 0 (0–2) .87 1 (0–4) 3 (1–7) <.001 0 (0–1) 0 (0–1) .95 
Duration of alarm, s 0 (0–0) 0 (0–24) <.001 0 (0–0) 0 (0–4) <.001 0 (0–0) 0 (0–0) .047 0 (0–0) 0 (0–26) <.001 
 0 (0–0) 1 (0–34) <.001 0 (0–0) 0 (0–19) .010 0 (0–2) 0 (0–2) .36 0 (0–0) 0 (0–23) .047 
 0 (0–0) 0 (0–14) <.001 2 (0–23) 0 (0–17) .65 8 (0–22) 10 (2–20) .49 0 (0–0) 0 (0–0) .86 
 0 (0–0) 0 (0–16) .009 0 (0–13) 0 (0–15) .96 2 (0–12) 17 (2–31) <.001 0 (0–10) 0 (0–19) .24 
Time in alarm, % 0 (0–0) 0 (0–11) <.001 0 (0–0) 0 (0–2) <.001 0 (0–0) 0 (0–0) .094 0 (0–0) 0 (0–13) <.001 
 0 (0–0) 0.3 (0–8) <.001 0 (0–0) 0 (0–5) .009 0 (0–0.6) 0 (0–0.5) .31 0 (0–0) 0 (0–5) .043 
 0 (0–0) 0 (0–4) <.001 0.56 (0–5) 0 (0–4) .96 2 (0–6) 3 (0.5–6) .37 0 (0–0) 0 (0–0) >.99 
 0 (0–0) 0 (0–4) .008 0 (0–4) 0 (0–5) .89 0.6 (0–3) 5 (0.6–8) <.001 0 (0–2) 0 (0–5) .23 

A P value <.05 is considered statistically significant. IQR, interquartile range.

FIGURE 2

Mean number of alarms due to LOSI by age.

FIGURE 2

Mean number of alarms due to LOSI by age.

Close modal
FIGURE 3

Mean number of episodes of LOSI without alarm by age.

FIGURE 3

Mean number of episodes of LOSI without alarm by age.

Close modal

This pilot study shows that continuously monitoring pulse oximetry with a partially wireless monitor without physical connection to a wall unit decreases episodes of LOSI generated by motion artifact in healthy pediatric subjects completing age-appropriate activities when compared with episodes of LOSI by using traditional wall-connected pulse oximetry. Use of partially wireless pulse oximetry results in lower number, shorter duration, and decreased percent time in alarm state due to LOSI compared with traditional wall-connected pulse oximetry at all activity levels tested in healthy children. To our knowledge, this is the first study investigating the relationship between implementation of partially wireless pulse oximetry on episodes of LOSI secondary to motion artifact during continuous pulse oximetry monitoring in children.

On the basis of the LOSI definition of a “Spo2 decrease >4%” from the subject’s baseline and episodes of “missing values in Spo2” corresponding to complete inability to pick up the patient signal, the data also show that the partially wireless pulse oximeter had significantly fewer number and shorter duration of episodes of LOSI not leading to alarm state compared with control pox. These findings indicate that in addition to decreasing false alarms secondary to LOSI from motion artifact, implementation of partially wireless pulse oximetry also decreases misleading device output (not triggering an alarm state) secondary to LOSI from motion artifact. The subgroup analysis indicates that partially wireless pulse oximetry may benefit the 6- to 10-year-old and 11 to 17-year-old patients more than the 1- to 5-year-old patients. This may be related to the fact under study conditions, youngest children were less vigorously active than older children. It will be interesting to see if this trend continues with in-hospital testing of patients.

Although there was clear statistical decrease in LOSI with the wireless pox compared with the control pox, the overall median number and duration of LOSI appear to be small in both groups. However, this pilot study was limited to ∼20 minutes of testing. In hospitalized children under continuous monitoring, this could translate to a significant decrease in clinical false alarms. This is not a trivial problem. There were 98 alarm-related events reported to the Joint Commission Sentinel Event database between January 2009 and June 2012, of which 80 resulted in death.2  The rates of alarm-related events are underreported and alarm-related deaths are estimated to be at least 10 times what the reported data show.14  In a study by Schondelmeyer, there were 220 813 audible alarms across all pediatric inpatient units during the 7-day study period, 30% of which were for low oxygen saturation and 33% were technical alarms including alarms for artifact and lead failure.15 

Schondelmeyer et al16  recently reported the frequency of monitor alarm rates at 5 children’s hospitals, finding a range of 42 to 351 alarms per monitored-patient day. In all 5 hospitals, the highest percent of total alarms were due to Spo2 being low.16  If extrapolated to a 24-hour period, our study indicates an even higher number of alarms per patient by using traditional pulse oximetry than was noted in the 5-hospital study. This likely is related to the intensive activity tested in our subjects, which was chosen to provoke the limits of LOSI in a pragmatic study design but could also be a factor of devices used, sampling frequency, or settings. Despite the differences in alarm rates between studies, it is clear that pulse oximetry is responsible for a large portion of audible alarms; therefore, the reduction in episodes of alarm state secondary to LOSI found with use of partially wireless pulse oximetry could assist with reduction of alarm fatigue hospital-wide, where use of pulse oximetry is routine.

There are several limitations of this study, including the small sample size and the fact that testing was limited to healthy pediatric subjects at normal oxygen saturation. The findings may vary in actual patients with varying disease pathology and potentially lower oxygen saturation values. Another limitation is the elimination of the infant age group due to slow enrollment because this age group may benefit from this technology. Also, for the purposes of this study, only the Bluetooth connectivity of the wireless pox was tested. LOSI due to “sensor off patient” was more common in control pox and occurred when pox probe fell off a child. During testing, the technician immediately replaced a displaced probe, which may not be representative of the inpatient response that may have a varying duration of time for probe replacement. Testing was also completed for a much shorter duration of time than in a hospitalized patient. In this study, we did not assess comfort or other measures of acceptability between the wireless pox and traditional wall-mounted pulse oximetry. Finally, because no quantitative definition for LOSI existed, proxies for LOSI were used for this study. Including analysis of perfusion index in future studies may assist with further validation of the proxies chosen for LOSI. Further testing needs to be completed to determine if these results are reproducible in the inpatient pediatric population over a longer duration of monitoring time by using wireless network connectivity in addition to the already tested Bluetooth connectivity.

It is known that reduction of false alarms is necessary to reduce alarm fatigue. The findings of this study indicate that implementation of partially wireless pulse oximetry in children may decrease episodes of LOSI and decrease false alarms from motion artifact.

The authors thank Mary Kasch (Clinical Research Assistant III) for her assistance with scheduling and conducting testing of subjects.

Dr Petersen conceptualized and designed the study, coordinated and collected data, and drafted the initial manuscript; Drs Hanson and Clarke participate in study design, supervised data collection, and reviewed and revised the manuscript; Dr Yan participated in study design, planned and conducted the analyses, and critically reviewed the manuscript; and all authors approved the final manuscript as submitted.

FUNDING: This project was internally funded by the Department of Pediatrics, Medical College of Wisconsin. Masimo (Irvine, CA) provided the required equipment via loan for this study. Masimo did not pay the investigators or institution to conduct this study. Masimo did not participate in data collection or analysis and did not restrict publication or influence the results of this study. Masimo was not involved in the authorship of this article.

1
Karnik
A
,
Bonafide
CP
.
A framework for reducing alarm fatigue on pediatric inpatient units
.
Hosp Pediatr
.
2015
;
5
(
3
):
160
163
2
Joint Commission on Accreditation of Healthcare Organizations [JCAHO]
.
R3_report: requirement, rationale, reference on alarm system safety. Joint Commission on Accreditation of Healthcare Organizations [JCAHO]. 2013. Available at: https://www.jointcommission.org/assets/1/18/R3_Report_Issue_5_12_2_13_Final.pdf. Accessed June 5, 2019
3
Schmid
F
,
Goepfert
MS
,
Franz
F
, et al
.
Reduction of clinically irrelevant alarms in patient monitoring by adaptive time delays
.
J Clin Monit Comput
.
2017
;
31
(
1
):
213
219
4
Goel
VV
,
Poole
SF
,
Longhurst
CA
, et al
.
Safety analysis of proposed data-driven physiologic alarm parameters for hospitalized children
.
J Hosp Med
.
2016
;
11
(
12
):
817
823
5
Schondelmeyer
AC
,
Brady
PW
,
Landrigan
CP
.
Alarm fatigue: clearing the air
.
J Hosp Med
.
2016
;
11
(
2
):
153
154
6
Lawless
ST
.
Crying wolf: false alarms in a pediatric intensive care unit
.
Crit Care Med
.
1994
;
22
(
6
):
981
985
7
Fouzas
S
,
Priftis
KN
,
Anthracopoulos
MB
.
Pulse oximetry in pediatric practice
.
Pediatrics
.
2011
;
128
(
4
):
740
752
8
Trivedi
NS
,
Ghouri
AF
,
Shah
NK
,
Lai
E
,
Barker
SJ
.
Effects of motion, ambient light, and hypoperfusion on pulse oximeter function
.
J Clin Anesth
.
1997
;
9
(
3
):
179
183
9
Barker
SJ
,
Shah
NK
.
The effects of motion on the performance of pulse oximeters in volunteers (revised publication)
.
Anesthesiology
.
1997
;
86
(
1
):
101
108
10
Clarke
G
.
Signal Quality Analysis in Pulse Oximetry: Modelling and Detection of Motion Artifact
. [master’s thesis].
Ottawa, ON
:
Ottawa-Carleton Institute for Biomedical Engineering, Carleton University
;
2015
11
Seixas
DM
,
Seixas
DM
,
Pereira
MC
,
Moreira
MM
,
Paschoal
IA
.
Oxygen desaturation in healthy subjects undergoing the incremental shuttle walk test
.
J Bras Pneumol
.
2013
;
39
(
4
):
440
446
12
Elder
JW
,
Baraff
SB
,
Gaschler
WN
,
Baraff
LJ
.
Pulse oxygen saturation values in a healthy school-aged population
.
Pediatr Emerg Care
.
2015
;
31
(
9
):
645
647
13
World Health Organization
.
Oxygen therapy for children: a manual for health workers
.
2016
. Available at: https://www.who.int/maternal_child_adolescent/documents/child-oxygen-therapy/en/. Accessed April 23, 2019
14
Sendelbach
S
,
Funk
M
.
Alarm fatigue: a patient safety concern
.
AACN Adv Crit Care
.
2013
;
24
(
4
):
378
386; quiz 387–388
15
Schondelmeyer
AC
,
Bonafide
CP
,
Goel
VV
, et al
.
The frequency of physiologic monitor alarms in a children’s hospital
.
J Hosp Med
.
2016
;
11
(
11
):
796
798
16
Schondelmeyer
AC
,
Brady
PW
,
Goel
VV
, et al
.
Physiologic monitor alarm rates at 5 children’s hospitals
.
J Hosp Med
.
2018
;
13
(
6
):
396
398

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: Equipment was provided by Masimo (Irvine, CA) via loan for this study; otherwise, all authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data