Skip to main content
NCBI home page
Journal of Child & Adolescent Trauma logoLink to Journal of Child & Adolescent Trauma
. 2020 Jun 10;14(2):165–175. doi: 10.1007/s40653-020-00313-1

Simultaneous Heart Rate Variability and Electroencephalographic Monitoring in Children in the Emergency Department

Juan A Piantino 1,✉, Amber Lin 2, Madison Luther 3, Luis D Centeno 4, Cydni N Williams 5, Craig D Newgard 6
PMCID: PMC8099962  PMID: 33986903

Abstract

Changes in heart rate variability (HRV) and electroencephalographic (EEG) background are promising tools for risk stratification and outcome prediction in children seen in the Emergency Department (ED). Novel monitoring technologies offer an opportunity for determining the clinical value of these physiologic variables, however, studies evaluating these measurements obtained in the Pediatric ED are sparse. The current study used a single center, prospective, observational cohort study of HRV and EEG as early predictors of outcome in children with acute trauma. ECG and HRV data were successfully collected in 167 subjects and simultaneous collection of ECG and EEG data using a wireless monitoring device was piloted in 17 patients with 15 patients having EEG data rated as appropriate for clinical interpretation. The mean time from ED arrival to ECG and EEG recording start was 7.5 (SD 11.6) and 34.5 (SD 15.5) minutes, respectively. The mean time required for EEG electrode placement was 9.3 min (SD 5.8 min). Results showed recording early HRV and EEG is feasible in children with acute injury seen in the ED. This study suggests that high consent rates are possible with the adequate research infrastructure and physiologic variables may offer an early, non-invasive marker for injury stratification and prognosis in children.

Electronic supplementary material

The online version of this article (10.1007/s40653-020-00313-1) contains supplementary material, which is available to authorized users.

Keywords: Heart rate variability, Electroencephalogram, Children, Multi-organ monitoring

Introduction

Emergency Department (ED) visits have steadily increased over the last decade. Each year, approximately 130 million patients visit EDs in the United States (Tang et al. 2010). Children account for 1/4 of these visits (“National Hospital Ambulatory Medical Care Survey: 2015 Emergency Department Summary Tables.,” 2015). Despite advances in Emergency Medicine Services for Children, several barriers to improving outcomes in this population remain, including difficulty with early recognition of serious injury, lack of individualized therapies, and lack of reliable prognostic measures. In recent years, two physiologic measures have been proposed as possible markers of injury severity, prognosis, and outcome: heart rate variability (HRV) and electroencephalographic (EEG) abnormalities (Henden et al. 2014; Payne et al. 2014). These measures address the integrity of two closely integrated systems that are crucial to normal physiologic function: the autonomic system and the central nervous system.

Decreased HRV, indicating autonomic dysregulation, has been associated with poor outcomes in adults with trauma, hemorrhagic shock, sepsis, congestive heart failure, and adrenal insufficiency (Batchinsky et al. 2007; Henden et al. 2014; Morris et al. 2007; Signorini et al. 2006; Werdan et al. 2009). EEG background abnormalities such as slowing of the faster EEG frequencies and voltage attenuation are indicators of cerebral dysfunction and they have been associated with increased mortality in children with acute brain injury (Abend and Dlugos 2007). Moreover, recent studies have shown a correlation between early HRV and EEG abnormalities in critically ill children who progressed to brain death (Piantino et al. 2019). While these physiologic markers have shown potential use in the identification and prognostication of critically ill children, acquiring these measures early in the clinical course presents several challenges. In addition to logistical and technical problems, prospective emergency care studies involving children often struggle with low consent rates (Amlie-Lefond et al. 2009). There are obstacles specific to the pediatric population that complicate the consent process. First, approaching a stressed parent who is unlikely to engage in conversations about research involving their child is a difficult task (Furyk et al. 2018). Second, children are considered a vulnerable population due to their inability to consent (Spriggs and Gillam 2008). Third, patients in the ED are also considered vulnerable given the urgency of their care (Castaneda-Guarderas et al. 2016). Fourth, children who arrive in the ED by ambulance or air transport are often alone, as parents are driving to the ED by private vehicle. These obstacles pose a major limitation to early data acquisition for research purposes in children.

In this study, our goal is to describe the methods, feasibility, and logistics of capturing early HRV and EEG data for children presenting to the ED. To demonstrate this methodology, we use a prospective cohort study of acutely injured children presenting to the ED. We discuss the consent process, infrastructure, devices and equipment, training, timing of data collection, and quality of data collected. We also discuss lessons learned through the design and implementation of this study. By disseminating this knowledge, we aim to provide other researchers with the tools necessary to advance the field of non-invasive acute monitoring.

Method

Participants

This study was conducted at Doernbecher Children’s Hospital Emergency Department in Portland, Oregon. Doernbecher is an accredited level I Pediatric trauma center with 145 beds, including a 20-bed PICU, over 65,000 inpatient admissions, and over 300 trauma patients presenting to the ED per year. Patients 1 month to 18 years of age with acute trauma were eligible to participate (Fig. 1). Trauma patients presenting to the Doernbecher ED trauma bay were recruited over a 24-month period between August 10, 2017 and August 28, 2019. All eligible patients underwent HRV monitoring. Patients unable to provide consent, pregnant women, and prisoners were excluded. In a subset of children, we piloted a protocol that includes simultaneous HRV and electroencephalogram recordings. Exclusion criteria specific to this subset of patients included: extensive head trauma that prevented placement of EEG leads, or patients who did not tolerate the EEG lead placement. EEG was performed on consecutive patients, until the target (n = 17) was met.

Fig. 1.

Fig. 1

Open in a new tab

Enrollment Flow Chart

Instruments/Measures

To assess EEG quality, recordings were divided into 10-s epochs and assessed on a 5 point Likert scale as follows: 1) uninterpretable (artifact present in the majority of the leads, and in more than 80% of the 10-s epochs, which significantly interferes with the interpretation of the study); 2) poor quality (artifact present in the majority of the EEG leads, and in 50–80% of the 10-s epochs); 3) adequate (EEG recording is adequate for clinical interpretation, artifact is present in 20–50% of the 10-s epochs and/or is limited to 1–2 leads, and does not interfere with clinical interpretation); 4) very good (artifact is present in less than 20% of the epochs and/or is limited to 1–2 leads, and does not interfere with clinical interpretation); 5) excellent (there is no artifact present). Research staff’s level of comfort operating the equipment was assessed with a 5-point Likert scale (ranging from “not comfortable at all” to “very comfortable”).

Procedures

The methods described here were implemented in a prospective, single-center cohort study of acutely injured children. The Institutional Review Board approved this protocol. In order to capture early physiologic data, the study was performed under deferred informed consent (described in detail below). Consent, data acquisition, and patient follow up were carried out by research assistants (RAs) who staff the ED and hospital 24 h per day, 7 days per week for trauma research through the Division of Trauma Surgery and Critical Care. This RA staff consists of 10 full time and 2 part-time clinical research assistants, plus 2 research interns (volunteers). We trained the research staff for 4 months prior to the study start date on all technical aspects of the study. These included recording raw ECG data from the bedside monitor (used to calculate HRV), placement of the EEG electrodes, and checking the quality of the acquired real-time data.

All members of the research team underwent a 2-h lecture by the study principal investigator (PI) on how to operate the ECG and EEG recording software, followed by a 2-h hands-on exercise under the PI’s direct supervision. We identified one research staff member to serve as the study lead. This person underwent an additional training, which included ways of testing for accurate placement of EEG electrodes according to established anatomical landmarks. The study lead had bi-weekly meetings with the study PI to discuss technical difficulties and other problems that arose during the study. For both ECG and EEG acquisition, each research staff member practiced on volunteers for a minimum of 5 times. They were allowed to continue practicing for as many times as needed to feel comfortable with the equipment, and until the EEG electrodes were accurately placed. Practice was supervised by the study lead.

Consent Process

Our protocol was deemed minimal risk by our Institutional Review Board (IRB) for the following reasons: a) ECG data is obtained as standard of care in the ED for all trauma patients; b) EEG recording presents no more than minimal risk to subjects; c) EEG is routinely performed in critically ill children without consent. In addition, the administration of EEG in a timely manner (prior to consent) was crucial to our study. For these reasons, the IRB provided authorization to conduct the study under deferred consent. Under this protocol, subjects retroactively consented to the use of data gathered as standard of care (ECG), as well as use of data gathered as part of procedures not considered standard of care (EEG) but considered minimal risk. Data collection started immediately after the patient arrived in the trauma bay, but consent was deferred until the patient was stabilized. The appropriate time to approach subjects for consent was determined by each RA on a case-by-case base. Consent was obtained either in the ED, the Pediatric ward, or the Pediatric Intensive Care Unit, depending on when the RA deemed appropriate to approach the family. Consent was obtained from a parent or legal authorized representative (LAR) if subjects were minors (12–17 years), or 18 years old but were unable to provide consent. Assent was obtained from children 12 to 18 years of age when possible. If subjects or their LARs decided not to participate in the study, the EEG and ECG files were erased from the computer.

Data Collection, Processing, and Analysis

Our goal was to obtain non-invasive physiologic data as early as possible after arrival in the ED. A research staff member was notified through the pediatric trauma paging activating system about all injured children coming to the ED by ambulance, typically 10 to 15 min prior to arrival. After receiving this notification, research staff placed a laptop computer (Dell Latitude 5290, Dell Inc., Round Rock, TX on an Ergotron StyleView SV10 stand [Ergotron Inc. St. Paul, MN]) and a wireless portable EEG recording device (Micro EEG, Biosignal Group, Acton, MA) in the ED trauma bay, and connected it to the bedside monitor (Phillips Intellivue MX800; Phillips, Bothel, WA) via an ethernet network cable. Setting up the monitoring unit took approximately 10 min. This unit was placed behind the monitor stand so as not to obstruct clinical care. As soon as the patient arrived in the trauma bay and was connected to the bedside monitor, the ECG signal was transmitted and stored on the laptop computer (Fig. 2). The portable EEG device was connected to the patient after completion of the primary and secondary examinations (described below).

Fig. 2.

Fig. 2

Open in a new tab

Overview of Data Acquisition and Analysis Protocol for Capturing HRV and EEG Data. *Note: Solid lines represent physical connections (electrodes, ethernet cables); dashed lines represent wireless connections

ECG Recording and HRV Calculation

ECG recording started immediately after the electrodes were placed on the patient’s chest. The signal was obtained while patients were lying supine on the ED gurney. Data was directly recorded from the bedside monitor using Medicollector Bedside (Medicollector, Inc. Boston, MA), at 250 Hz sampling frequency. We recorded for a minimum of 5 min. After the data was recorded, the laptop was disconnected from the bedside monitor, and the first 5 min of ECG monitoring were securely uploaded via WiFi and saved in the hospital server for further analysis. Each patient was assigned a study ID, and all stored data was otherwise deidentified of protected health information.

According to our institution’s Information Technology Office policies, raw ECG data were stored in an encrypted, password-protected database. HRV extraction based on raw ECG data was performed the following day and took approximately 5 min per patient. We analyzed the ECG data using Kubios (Biosignal Medical Group, Kuopio, Finland) software (Seppala et al. 2014; Tarvainen et al. 2014). The techniques and methods used to analyze HRV, in accordance with published guidelines (“Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology,” 1996), are briefly described below and depicted in Fig. 3 (see below). A more detailed description can be found elsewhere (Tarvainen et al. 2014). First, R-wave peaks were detected using an automated QRS detection algorithm. The detection process was manually inspected for potential errors. Once the R to R intervals were obtained, linear HRV parameters in the time and frequency-domain, as well as non-linear indexes (including Poincaré plots and approximate entropy [ApEn]) were performed. Time domain parameters included the mean value of the NN intervals, the standard deviation of the NN intervals (SDNN), and the root mean square of successive NN interval differences (rMSSD). Frequency-domain analysis was performed with the non-parametric method of fast Fourier transformation. The resulting heart rate spectra were quantified under two components: low-frequency (LF: 0.04–0.15 Hz) and high frequency (HF: 0.15–0.4 Hz) parameters, as well as LF/HF ratio. Values with skewed distribution were log-transformed.

Fig. 3.

Fig. 3

Open in a new tab

Examples of HRV Data Obtained in the Emergency Department. *Notes: A: raw ECG data imported from the Philips monitor, the red crosses indicate the automated R wave identification. B and B′: RR interval analysis in two different patients with acute trauma, GCS of 15 on arrival (B), GCS of 3 on arrival (B′). Red arrows indicate variation of R-R interval with respiration. C and C′: frequency domain HRV analysis on the same two patients, using fast Fourier transform (FFT). The colors indicate the very low frequency power (grey), low frequency power (red); and high frequency power (green). Note the decrease in power, indicating low variability, in the subject with low GCS (C′)

EEG Recording

We piloted the use of simultaneous ECG and EEG from a single device in 17 patients. Monitoring was performed using a wireless Micro-EEG device, shown in Fig. 4 (Micro-EEG, Biosignal Inc., Acton, MA), as soon as permitted by the patient’s status. A study staff member placed gel EEG electrodes (HydroDot® biosensors, Biosignal Inc., Acton, MA) using an FDA-approved, semi-disposable, 22 channel electrode headset (EzeNet®, Biosignal Group, Acton, MA) according to the international 10–20 configuration. The EzeNet® is a semi-disposable EEG electrode cap that provides full EEG coverage for both pediatric and adults patents. The EzeNet® is connected to the Micro EEG device, a portable, wireless EEG transmitted which can be secured to the patient’s head or wrist. Electrode impedance threshold was kept below 5 KOhm, sensitivity 7 microvolts, notch filter on at 60 Hz, low frequency filter at 3 s, high frequency filter at 30 Hz.. Study staff members were trained to check electrode impedance and quality of the EEG recording (Figs. 5 and 6). EEG recordings continued for 10 min in the ED trauma bay. After finishing the recording, EEG files were saved to the secure hospital server via WiFi with a generic (.edf) file extension, which can be opened by most EEG reviewing software.

Fig. 4.

Fig. 4

Open in a new tab

Wireless EEG Device Used to Obtain EEG Data in Patients in the Emergency Department. *Note: A. The EzeNet®semi-disposable EEG electrode cap. B.Micro EEG device. (Images published with permission from Biosignal, Inc.)

Fig. 5.

Fig. 5

Open in a new tab

EEG Tracing Obtained from a 2-year-old Boy with Acute Traumatic Brain Injury (GCS 11). *Note: red arrow shows the presence of a slow posterior dominant rhythm

Fig. 6.

Fig. 6

Open in a new tab

Example of an EEG Recording with Electrode Artifact Obtained on a 10-year old with Acute Trauma (GCS 15). Note: red arrows indicate motion artifact in P3 and Pz electrodes. Red arrowheads indicate an 8 Hz posterior dominant rhythm

All the EEG recordings were number-coded, and interpreted off-line by a board certified Pediatric Epileptologist who was blind to clinical information about the subject, using Persyst 13 software (Persyst Inc. Prescott, AZ). Of note, although Micro-EEG devices allow real-time streaming of EEG data for immediate interpretation in the clinical setting, in our study EEGs were analyzed the following day. The analysis and interpretation of a 10-min recording took approximately 20 min.

Outcomes

The primary outcomes for the study included hospital length of stay, and Intensive Care Unit length of stay. In addition, Pediatric Functional Status Score and Glasgow Outcome Score were obtained by phone or chart review at 1 and 6 months.

Data Analysis

The target sample size for this study was 150 subjects. We analyzed the data using descriptive statistics (mean and median as appropriate based on distribution of continuous data, and proportions for categorical). Statistical analyses were performed using Stata/MP 15 (StataCorp LP, College Station, TX).

Results

Enrollment

We screened 314 children in the ED for eligibility during the enrollment period (Fig. 1). Nine patients did not meet inclusion criteria, 50 patients were discharged before the research staff was able to reach parents for consent, and 1 died before parents were approached for consent. Of the 254 patients who were approached for consent, 238 agreed to participate (consent rate 93.7%).

Data Collection and Quality – HRV

Seventy-one patients who consented to participate (29.8%) were excluded from the HRV analysis. Forty-seven of those (19.7%) due to technical problems obtaining the recording, and 24 (10%) due to less than the minimum 5-min recording required by the study. Technical problems were not related to the software, and included motion artifact that degraded the ECG quality, operator errors (i.e. selecting the wrong channels to record), and operator unable to hook up the computer on time due to patients arriving earlier than expected. The final cohort consisted of 167 patients. Of those, 54 were between the ages of 12 and 17, and therefore eligible to provide assent. Four (7%) of those patients provided assent. The main reason for not obtaining assent was altered mental status. Follow-up rates are 93.2% and 77.5% at 1 and 6 months, respectively.

Both time of arrival to the ED, and stat time of ECG recording were available in 157 subjects. Using Medicollector, the mean time from arrival to the ED to ECG recording start was 7.5 min (SD 11.6 min). All 10 RAs participated in the collection of ECG data from the bedside monitors. After 3 supervised trials, all felt comfortable connecting the computer to the monitor, and running the software.

Data Collection and Quality – EEG

A protocol aimed at establishing feasibility of simultaneous HRV and EEG monitoring was piloted in 17 patients. The median age was 8.8 years (range 1.8 to 16.8). Five RAs participated in the pilot EEG study. EEG data was collected for at least 10 min in order to allow adequate assessment of the quality of the recording. In some cases, two RAs applied electrodes on the same subject. After 3 supervised trials, all but one RA felt comfortable or very comfortable operating the EEG device and placing the electrodes. Fifteen out of 17 EEG recordings were rated 3 (EEG recording is adequate for clinical interpretation, artifact is present in 20–50% of the 10-s epochs and/or is limited to 1–2 leads, and does not interfere with clinical interpretation) or above by a board certified Pediatric Epileptologist. Six EEG recordings were rated as “very good” to “excellent”, meaning that they had minimal to no artifact. Figure 5 shows an EEG tracing with no electrode artifact. Figure 6 shows an EEG which was graded a 3: EEG recording is adequate for clinical interpretation, artifact is present in 20–50% of the 10-s epochs and/or is limited to 1–2 leads, and does not interfere with clinical interpretation. Using the Micro EEG device, the mean time from arrival to the trauma bay to EEG recording start was 34.5 min (SD 15.5 min). The mean time required for electrode placement was 9.3 min (SD 5.8 min).

Discussion

We report the methodology for capturing early non-invasive physiologic variables in acutely injured children presenting to the ED, including quality assessment of the data. We conclude that capture of high-quality, early multi-organ, noninvasive physiologic data in children is feasible shortly after ED arrival, with a high consent rate using a deferred consent model.

In recent years, there has been increased interest in the utilization of noninvasive physiologic markers such as HRV or EEG changes as early, noninvasive indicators of injury severity and outcome both in the pediatric and adult population (Henden et al. 2014; King et al. 2009; Payne et al. 2014; Piantino et al. 2019; Ryan et al. 2011). This has been fueled in part by the appearance of novel technologies which allow researchers to capture and analyze continuous physiologic data in these patients (Bradley et al. 2012; Ibrahim et al. 2016; Kuppermann et al. 2009; Vinecore et al. 2007; Zehtabchi et al. 2014).

These methodologies, although promising, require demonstration of feasibility in real-life situations. While articles detailing the methodology of multi-organ variability recording in the adult and pediatric intensive care unit have been published (Bradley et al. 2012; Vinecore et al. 2007), there is sparse literature describing the informatics, patient tracking, data processing, logistics, and other analysis for observational, prospective pediatric emergency care studies. The discussion below focuses on our approach to common problems encountered in the field, and the lessons we learned in the process of successfully completing our observational study.

Lesson 1: Deferred Consent Can Be Used to Gather Physiologic Data Non-invasively in the Pediatric ED. this Strategy May Lead to High Enrollment and Consent Rates

In this study we used a team of RAs dedicated to enrolling subjects on a 24/7 basis and trained in the specifics of consenting for pediatric studies. The observational, minimal risk nature of our study and the need for data immediately after ED arrival allowed us to obtain a deferred consent exemption from our institution’s IRB. While research under exception from informed consent has been performed in adults (MacKay et al. 2015), its applicability in children has been less explored. Although debated in the scientific community as it appears to compromise the principle of autonomy (Furyk et al. 2018; MacKay et al. 2015; “National Commission for the Protection of Human Subjects of Biolchemical and Behavioral Research. Belmont report: ethical principles and guidelines for the protection of human subjects of research.,” April 18, 1979), deferred consent has been proposed as a viable approach to performing research in the Pediatric ED (Brierley and Larcher 2011).

Despite its benefits, the use of deferred consent carries an inherent risk of exposing vulnerable populations to unnecessary procedures. This risk is exacerbated by the lack of a clear definition of “minimal risk”. International guidelines establish that deferred consent can be undertaken if the research cannot be delayed, risk is low, there is potential benefit to the child, the research has merit, there is no reason to suspect that the parents would not vice consent, and consent is obtained as soon as possible (World Medical Association 2013). However, important definitions such as the extent of “minimal risk” are left to the discretion of each individual institution, leading to discrepancies in what the type of protocols children are or are not allowed to participate. In our study, the placement of EEG leads was deemed “minimal risk”. Under this protocol, 238 out of 254 parents approached for consent agreed to participate. This suggests that observational research in the Pediatric ED under deferred consent is feasible and that high consent rates are achievable. Of note, 50 subjects were excluded because parents left the hospital before being approached for consent. Although not attempted in this study, phone consent is a potential solution to this problem.

Lesson 2: Check the Quality of the Physiologic Data Frequently. Even the Simplest Data Acquisition Protocol Is Subject to Operator Error

In addition to the barriers inherent to the consent process, capture of novel physiologic markers in the pediatric population requires collection of high quality, artifact-free data. In recent years, new HRV and EEG acquisition devices and software have improved our ability to record data in the acute setting (Grant et al. 2014; Parvizi et al. 2018; Smith et al. 2015). Despite their innovative and user-friendly platforms, these devices still require training in order to achieve adequate data collection. In our study, we devoted 4 months for training of our research team prior to initiation of enrollment. Even under this extensive training protocol, 47 out of 238 eligible patients had to be excluded due to technical problems with data acquisition from the bedside monitor. Rather than software malfunction, these problems were related to operator errors. Frequent data monitoring is one way to minimize this loss of valuable data.

Lesson 3: Identify One RA as the Study Lead. They Will Be Responsible to Train the Other RAs and Keep their Skills Updated (“train the trainer” Model)

We identified research staff turnaround as the main contributor to operator errors and increased supervision of new staff by the study lead. This approach eventually solved the problem. Also, much of our efforts leading to the study start date were focused on finding user-friendly equipment. ECG acquisition devices for HRV analysis range from single-patient bedside monitoring units (Szatala and Young 2019; Vinecore et al. 2007) to larger, multi-bed monitoring platforms (Megjhani et al. 2018). For our study, we chose a portable, single-bed acquisition device, which was connected via an ethernet cable to the Phillips bedside monitor. Post-processing of the ECG data was done with Kubios software. Of the software available for testing, we found Kubios, which uses a Matlab (MathWorks, Inc., Natick, MA) platform, to be the easiest to operate. The use of similar equipment has been reported in critically ill children in the ICU (Vinecore et al. 2007). However, the ICU and the ED are significantly different environments, hence the need for ED-specific studies.

Lesson 4: EEGs Can Be Performed in the ED in the Acute Setting, Even in Children As Young as 1 Year

Performance of a conventional EEG in the pediatric ED has its own set of challenges. Though most large pediatric tertiary care centers have 24/7 EEG technician availability, they are often unavailable to perform EEGs immediately in the ED (Sanchez et al. 2013). Moreover, placing conventional EEG electrodes requires 40 min to one hour. Conventional EEG devices, even those designed for ICU monitoring, have a large footprint that may compromise the trauma bay workflow and may not be compatible with magnetic-resonance imaging should the patient need urgent imaging. Last, acutely injured children are often agitated and crying, which complicates the placement of conventional EEG electrodes. In recent years, several devices have been developed to address some of these limitations, both in adults and children (Omurtag et al. 2012; Parvizi et al. 2018). The majority of these devices consist of a headband or cap which can be quickly applied to the patient’s head. Electrodes are then placed in pre-specified slots or are incorporated in the headband itself. These devices differ in the type of electrodes they utilize, availability of pediatric headbands, the head surface covered (full 20–20 versus limited montage) and use of disposable or re-usable bands. Most devices transmit to a unit which is connected wirelessly to the hospital server either directly or via a portable computer. From there, readers can visualize the EEG in real-time.

For our study, we decided to use the Micro-EEG device because the availability of different pediatric head sizes, which makes it more usable in younger patients, the availability to provide a full-array of EEG electrodes, the availability to use a video camera (attached to the laptop computer), and its small footprint (Omurtag et al. 2012). This device transmits directly to our portable computer, and allows for simultaneous EEG and ECG recording. One of the benefits of collecting data directly from the bedside monitors and EEG devices is the ability to compute HRV and EEG power spectral analysis in real time, whereas other platforms are appropriate for retrospective analysis only. Larger data acquisition platforms such as Bedmaster® allow real-time data acquisition, as well as integration with Philips, General Electric, and other ambulatory devices. These platforms; however, are costly and usually reserved for larger, multi-bed monitoring efforts.

One of the disadvantages of utilizing portable data acquisition systems is that they require a bedside computer, and communication interface protocols to access data through external ports on the monitors (Vinecore et al. 2007). With the setup used in our study, we found this to be a minimal impediment, and problems with the monitoring unit footprint were not reported, even in the often-crowded trauma bay. Another disadvantage is the need for one portable computer for each monitored bed. Multi-bed monitoring platforms have been successfully used in the ICU (Bradley et al. 2012). However, we did not have the need for a multi-bed monitoring platform because we seldom had more than one child in the trauma bay at a time.

Lesson 5: Stakeholder Engagement to Establish a Clear Protocol and Minimize Disruptions to Clinical Workflow Is Essential

Placement of hardware on an unstable patient for research purposes without obstructing patient care is difficult. The fast-paced workflow of a trauma bay means that monitoring equipment must be ready to record upon patient arrival and have minimal footprint. The trauma surgery team, who leads evaluation of all trauma patients in the ED, was involved early on in the design and implementation of our study. Their feedback regarding timing, equipment location, and consent process was incorporated to our protocol, which allowed us to have minimal disruption of their workflow.

Limitations

This study has several limitations. First, we report on one logistical plan for data acquisition. As stated above, the system was chosen based on our technical and logistical needs. It is possible that other devices and acquisition protocols perform equally well or better than the one described here. Each platform and strategy must adapt to the scientific question being addressed. Second, despite our efforts to enroll early and cover the ED at all times, 50 patients left the hospital before consent was obtained. This is a problem inherent to performing research in the ED and must be taken into account when projecting enrollment rates in observational studies. Third, we report a small number of EEGs performed in this study. The purpose of obtaining EEGs was to establish the feasibility of multi-organ monitoring shortly after ED arrival. We did not compare the quality of EEGs obtained through different acquisition devices or compare to standard in-hospital EEG acquisition (which may be performed hours to days later, or not at all). It is possible that devices that are easier to deploy than the full montage EEG cap would provide similar information. A true comparison would also require a “gold standard” such as conventional EEG, which is difficult to obtain in the ED. Fourth, while this study demonstrated feasibility of consenting parents for pediatric research, it relied heavily on a robust research infrastructure at a tertiary academic institution. Its applicability to smaller EDs should be studied in more detail. Continuous multi-organ physiologic monitoring in acutely injured children has been the subject of extensive recent research, and has the potential to directly impact patient care, assist research in Emergency Care, and foster research in novel monitoring technology. Future research will be focused on the applicability of real-time HRV and EEG monitoring to improve risk stratification and prognostication in acutely injured children. This research has the potential to significantly impact the care of children in the ED, extend to the Pediatric ICU.

Conclusions

Capturing high-quality non-invasive physiologic variables early in the clinical course among children presenting to the ED is feasible. Deferred consent may aid with high enrollment and consent rates. We describe the processes and techniques used to collect HRV and EEG data from the ED bedside. A robust research infrastructure, pre-implementation stakeholder engagement, and careful selection of data acquisition platforms are needed to perform large prospective observational studies needed in this field.

Electronic supplementary material

ESM 1 (4.4MB, png)

(PNG 4528 kb)

High Resolution image (TIF 8493 kb) (8.3MB, tif)

Acknowledgments

Dr. Piantino’s institution received funding from the National Heart, Lung, and Blood Institute (NHLBI) K12HL133115. Drs. Piantino and Newgard, and Ms. Lin received support for article research from the National Institutes of Health (NIH). Dr. Williams’ institution received funding from, and she received support for article research from the Agency for Healthcare Research and Quality K12HS022981. Dr. Newgard’s institution received funding from NIH/NHLBI grant number K12HL133115.

Compliance with Ethical Standards

Disclosure of Interest

All authors declare they have no disclosures of interest to report.

Ethical Standards and Informed Consent

all procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation [institutional and national] and with the Helsinki Declaration of 1975, as revised in 2000. Informed consent was obtained from all patients for being included in the study.”

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Abend NS, Dlugos DJ. Nonconvulsive status epilepticus in a pediatric intensive care unit. Pediatric Neurology. 2007;37(3):165–170. doi: 10.1016/j.pediatrneurol.2007.05.012. [DOI] [PubMed] [Google Scholar]
  2. Amlie-Lefond C, Chan AK, Kirton A, deVeber G, Hovinga CA, Ichord R, et al. Thrombolysis in acute childhood stroke: Design and challenges of the thrombolysis in pediatric stroke clinical trial. Neuroepidemiology. 2009;32(4):279–286. doi: 10.1159/000203076. [DOI] [PubMed] [Google Scholar]
  3. Batchinsky AI, Cooke WH, Kuusela T, Cancio LC. Loss of complexity characterizes the heart rate response to experimental hemorrhagic shock in swine. Critical Care Medicine. 2007;35(2):519–525. doi: 10.1097/01.CCM.0000254065.44990.77. [DOI] [PubMed] [Google Scholar]
  4. Bradley B, Green GC, Batkin I, Seely AJ. Feasibility of continuous multiorgan variability analysis in the intensive care unit. J Crit Care. 2012;27(2):218.e219–218.e220. doi: 10.1016/j.jcrc.2011.09.009. [DOI] [PubMed] [Google Scholar]
  5. Brierley J, Larcher V. Emergency research in children: Options for ethical recruitment. Journal of Medical Ethics. 2011;37(7):429–432. doi: 10.1136/jme.2010.040667. [DOI] [PubMed] [Google Scholar]
  6. Castaneda-Guarderas A, Glassberg J, Grudzen CR, Ngai KM, Samuels-Kalow ME, Shelton E, et al. Shared decision making with vulnerable populations in the emergency department. Academic Emergency Medicine. 2016;23(12):1410–1416. doi: 10.1111/acem.13134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Furyk J, McBain-Rigg K, Renison B, Watt K, Franklin R, Emeto TI, et al. A comprehensive systematic review of stakeholder attitudes to alternatives to prospective informed consent in paediatric acute care research. BMC Medical Ethics. 2018;19(1):89. doi: 10.1186/s12910-018-0327-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Grant AC, Abdel-Baki SG, Omurtag A, Sinert R, Chari G, Malhotra S, et al. Diagnostic accuracy of microEEG: A miniature, wireless EEG device. Epilepsy & Behavior. 2014;34:81–85. doi: 10.1016/j.yebeh.2014.03.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Henden PL, Sondergaard S, Rydenhag B, Reinsfelt B, Ricksten SE, Aneman A. Can baroreflex sensitivity and heart rate variability predict late neurological outcome in patients with traumatic brain injury? Journal of Neurosurgical Anesthesiology. 2014;26(1):50–59. doi: 10.1097/ANA.0b013e3182a47b62. [DOI] [PubMed] [Google Scholar]
  10. Ibrahim ZH, Chari G, Abdel Baki S, Bronshtein V, Kim MR, Weedon J, Cracco J, Aranda JV. Wireless multichannel electroencephalography in the newborn. J Neonatal Perinatal Med. 2016;9(4):341–348. doi: 10.3233/npm-161643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. King DR, Ogilvie MP, Pereira BM, Chang Y, Manning RJ, Conner JA, et al. Heart rate variability as a triage tool in patients with trauma during prehospital helicopter transport. The Journal of Trauma. 2009;67(3):436–440. doi: 10.1097/TA.0b013e3181ad67de. [DOI] [PubMed] [Google Scholar]
  12. Kuppermann N, Holmes JF, Dayan PS, Hoyle JD, Jr, Atabaki SM, Holubkov R, et al. Identification of children at very low risk of clinically-important brain injuries after head trauma: A prospective cohort study. Lancet. 2009;374(9696):1160–1170. doi: 10.1016/S0140-6736(09)61558-0. [DOI] [PubMed] [Google Scholar]
  13. MacKay CR, Torguson R, Waksman R. Delayed consent: Will there be a shift in approach for US primary percutaneous coronary intervention trials? Lancet. 2015;386(9994):714–716. doi: 10.1016/s0140-6736(15)60077-0. [DOI] [PubMed] [Google Scholar]
  14. Megjhani M, Terilli K, Martin A, Velazquez A, Claassen J, Roh D, et al. Deriving the PRx and CPPopt from 0.2-Hz data: Establishing generalizability to Bedmaster users. Acta Neurochirurgica. Supplement. 2018;126:179–182. doi: 10.1007/978-3-319-65798-1_37. [DOI] [PubMed] [Google Scholar]
  15. Morris JA, Norris PR, Waitman LR, Ozdas A, Guillamondegui OD, Jenkins JM. Adrenal insufficiency, heart rate variability, and complex biologic systems: A study of 1,871 critically ill trauma patients. Journal of the American College of Surgeons. 2007;204(5):885–892. doi: 10.1016/j.jamcollsurg.2007.01.019. [DOI] [PubMed] [Google Scholar]
  16. Omurtag A, Baki SG, Chari G, Cracco RQ, Zehtabchi S, Fenton AA, Grant AC. Technical and clinical analysis of microEEG: A miniature wireless EEG device designed to record high-quality EEG in the emergency department. International Journal of Emergency Medicine. 2012;5(1):35. doi: 10.1186/1865-1380-5-35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Parvizi J, Gururangan K, Razavi B, Chafe C. Detecting silent seizures by their sound. Epilepsia. 2018;59(4):877–884. doi: 10.1111/epi.14043. [DOI] [PubMed] [Google Scholar]
  18. Payne ET, Zhao XY, Frndova H, McBain K, Sharma R, Hutchison JS, Hahn CD. Seizure burden is independently associated with short term outcome in critically ill children. Brain. 2014;137(Pt 5):1429–1438. doi: 10.1093/brain/awu042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Piantino JA, Lin A, Crowder D, Williams CN, Perez-Alday E, Tereshchenko LG, Newgard CD. Early heart rate variability and electroencephalographic abnormalities in acutely brain-injured children who Progress to brain death. Pediatric Critical Care Medicine. 2019;20(1):38–46. doi: 10.1097/pcc.0000000000001759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Ryan ML, Ogilvie MP, Pereira BM, Gomez-Rodriguez JC, Manning RJ, Vargas PA, et al. Heart rate variability is an independent predictor of morbidity and mortality in hemodynamically stable trauma patients. The Journal of Trauma. 2011;70(6):1371–1380. doi: 10.1097/TA.0b013e31821858e6. [DOI] [PubMed] [Google Scholar]
  21. Sanchez SM, Carpenter J, Chapman KE, Dlugos DJ, Gallentine WB, Giza CC, et al. Pediatric ICU EEG monitoring: Current resources and practice in the United States and Canada. Journal of Clinical Neurophysiology. 2013;30(2):156–160. doi: 10.1097/WNP.0b013e31827eda27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Seppala S, Laitinen T, Tarvainen MP, Tompuri T, Veijalainen A, Savonen K, Lakka T. Normal values for heart rate variability parameters in children 6-8 years of age: The PANIC study. Clinical Physiology and Functional Imaging. 2014;34(4):290–296. doi: 10.1111/cpf.12096. [DOI] [PubMed] [Google Scholar]
  23. Signorini MG, Ferrario M, Marchetti M, Marseglia A. Nonlinear analysis of heart rate variability signal for the characterization of cardiac heart failure patients. Conference Proceedings: Annual International Conference of the IEEE Engineering in Medicine and Biology Society. 2006;1:3431–3434. doi: 10.1109/IEMBS.2006.259744. [DOI] [PubMed] [Google Scholar]
  24. Smith VC, Kelty-Stephen D, Qureshi Ahmad M, Mao W, Cakert K, Osborne J, Paydarfar D. Stochastic resonance effects on apnea, Bradycardia, and oxygenation: A randomized controlled trial. Pediatrics. 2015;136(6):e1561–e1568. doi: 10.1542/peds.2015-1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Spriggs MP, Gillam LH. Consent in paediatric research: An evaluation of the guidance provided in the 2007 NHMRC national statement on ethical conduct in human research. The Medical Journal of Australia. 2008;188(6):360–362. doi: 10.5694/j.1326-5377.2008.tb01655.x. [DOI] [PubMed] [Google Scholar]
  26. Szatala A, Young B. Implementation of a data acquisition and integration device in the neurologic intensive care unit. AACN Advanced Critical Care. 2019;30(1):40–47. doi: 10.4037/aacnacc2019188. [DOI] [PubMed] [Google Scholar]
  27. Tang N, Stein J, Hsia RY, Maselli JH, Gonzales R. Trends and characteristics of US emergency department visits, 1997-2007. JAMA. 2010;304(6):664–670. doi: 10.1001/jama.2010.1112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Tarvainen MP, Niskanen JP, Lipponen JA, Ranta-Aho PO, Karjalainen PA. Kubios HRV--heart rate variability analysis software. Computer Methods and Programs in Biomedicine. 2014;113(1):210–220. doi: 10.1016/j.cmpb.2013.07.024. [DOI] [PubMed] [Google Scholar]
  29. Vinecore K, Aboy M, McNames J, Phillips C, Agbeko R, Peters M, et al. Design and implementation of a portable physiologic data acquisition system. Pediatric Critical Care Medicine. 2007;8(6):563–569. doi: 10.1097/01.pcc.0000288715.66726.64. [DOI] [PubMed] [Google Scholar]
  30. Werdan, K., Schmidt, H., Ebelt, H., Zorn-Pauly, K., Koidl, B., Hoke, R. S., , . . . Müller-Werdan, U. (2009). Impaired regulation of cardiac function in sepsis, SIRS, and MODS. Canadian Journal of Physiology and Pharmacology, 87(4), 266–274. 10.1139/Y09-012. [DOI] [PubMed]
  31. World Medical Association Declaration of Helsinki: Ethical principles for medical research involving human subjects. JAMA. 2013;310(20):2191–2194. doi: 10.1001/jama.2013.281053. [DOI] [PubMed] [Google Scholar]
  32. Zehtabchi S, Abdel Baki SG, Omurtag A, Sinert R, Chari G, Roodsari GS, et al. Effect of microEEG on clinical management and outcomes of emergency department patients with altered mental status: A randomized controlled trial. Academic Emergency Medicine. 2014;21(3):283–291. doi: 10.1111/acem.12324. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ESM 1 (4.4MB, png)

(PNG 4528 kb)

High Resolution image (TIF 8493 kb) (8.3MB, tif)

Articles from Journal of Child & Adolescent Trauma are provided here courtesy of Springer

RESOURCES