Real-time State of Vigilance Monitor for the Neonatal Intensive Care Unit

The goal of this observational study is to collect data to develop a complete package (hardware, user interface software and algorithms) that can monitor sleep-wake stages in neonates. Real-time EEG data will be used to develop and refine the prototype monitor's ability to provide direct real-time information about sleep-wake state. The study design includes multiple iterative training/testing stages to refine the prototype. The study is divided into multiple sub-aims conducted in parallel: data acquisition, algorithm development (including comparison between gold-standard polysomnogram vs. novel algorithm markings of sleep-stages), and graphical user interface software development. The data acquisition and algorithm development are iterative and linked, such that the prototype algorithm from one iteration will be deployed real-time during the next iteration of data acquisition. This allows verification that the algorithm can perform real-time and provides prospective testing data, which is later folded into the training data for the next iteration, for verification and validation of the system.

Study Overview

• Status: Recruiting
• Conditions: Neonatal Disease; Sleep Disturbance in Infancy (Disorder)
• Intervention / Treatment: Other: Novel Real-time Neonatal Sleep Stage Detection Algorithm

Detailed Description

Use Case: Disruption of sleep is a common experience of hospitalized patients of all ages, especially if they are in an intensive care unit (ICU); infants in the NICU often stay weeks to months. The quality of neonatal sleep is strongly associated with later neurodevelopment. Abnormal sleep quality in neonates is associated with less attention orienting at 4 months, increased distractibility, decreased developmental function at 12-24 months, and lower emotional regulation and cognitive development at age 5. Yet disruptions to sleep are frequent: the study team has found that NICU neonates are handled by staff with a median interval of 2.3 min. Handling occurred across all sleep-wake stages and frequently resulted in arousals, awakenings, and respiratory events.

The study will employ multiple levels of noise reduction and signal quality assessment, tailored to the specific needs of the algorithm and to the NICU environment. The final algorithm will include four sleep-wake stages (awake, quiet/non-REM sleep, active/REM sleep, indeterminate/transitional sleep). The complete algorithm is intended to do what no existing algorithm does, as it uniquely combines the assessment of data quality, ability to run real-time on un-curated data, and identification of four sleep-wake stages. In this observational study, validation of the monitor "read-out" in comparison to the gold standard polysomnogram happens offline, after the completion of data collection from the subject.

Because this is an observational study to develop and validate a new physiologic monitoring modality (the bedside sleep monitor), there is no subject intervention. No change in care for these study subjects is envisioned to result from use of this prototype monitor. It is hoped that the proposed technology will create a unique solution that will one day be deployed in the NICU.

Potential Future Research Use of Monitor: The study team hypothesizes that adjusting the timing of NICU patient contact to avoid multiple sleep disruptions might improve sleep in the NICU, which could then lead to better neurodevelopmental outcomes. However, the efficacy of this approach would need to be proved in randomized controlled trials. The conduct of such a future randomized clinical trial would be facilitated by availability of a real-time, objective monitor of sleep-wake states like what the study team proposes to develop. However, the present observational study serves as an initial stage of research that would lead up to that kind of trial.

• Study Type: Observational
• Enrollment (Estimated): 32

Contacts and Locations

• Study Contact: Name: Stephanie Rau, BS; Phone Number: 734-936-4179; Email: shatchew@med.umich.edu

Study Locations

• United States
  • Michigan
    • Ann Arbor, Michigan, United States, 48109
      • Recruiting
      • University of Michigan
      • Contact: John Barks, MD; Phone Number: 734-763-4109; Email: jbarks@umich.edu
      • Principal Investigator: John Barks, MD
      • Principal Investigator: Stephen Gliske, PhD

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: 1 day to 1 year (Child)
• Accepts Healthy Volunteers: No

• Sampling Method: Probability Sample

Study Population

Newborns admitted to the Newborn ICU in C&W Mott Hospital

Description

Inclusion Criteria:

• Admitted to the Newborn ICU in C&W Mott Hospital
• >/= 30 weeks gestational age at the time of birth (>/= 33 weeks post-conceptional age at enrollment)

Exclusion Criteria:

• Any diagnosis, patient care, or anticipated patient care that is likely to interfere with the 12 hour recording or would make the recording dangerous to the participant

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
NICU Cohort; Participants will undergo a standard polysomnogram	Other: Novel Real-time Neonatal Sleep Stage Detection Algorithm; Comparison of Novel Real-time Neonatal Sleep Stage Detection Algorithm to gold standard concurrent polysomnogram.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Negative predictive value (NPV) for detection of sleep (any stage)	Negative predictive value of monitor compared to gold-standard polysomnogram	12 hours

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
NPV for detection of REM sleep	Negative predictive value of monitor compared to gold-standard polysomnogram	12 hours
NPV for detection of non-REM sleep	Negative predictive value of monitor compared to gold-standard polysomnogram	12 hours
Positive predictive value (PPV) of Sleep	Positive predictive value of monitor compared to gold-standard polysomnogram	12 hours
Positive predictive value (PPV) of REM sleep	Positive predictive value of monitor compared to gold-standard polysomnogram	12 hours
Positive predictive value (PPV) of non-REM sleep	Positive predictive value of monitor compared to gold-standard polysomnogram	12 hours

Collaborators and Investigators

• Sponsor: University of Michigan
• Collaborators: National Heart, Lung, and Blood Institute (NHLBI); University of Nebraska
• Investigators: Principal Investigator: John Barks, MD, University of Michigan; Principal Investigator: Stephen Gliske, PhD, University of Nebraska

Study record dates

Study Major Dates

• Study Start (Actual): September 1, 2021
• Primary Completion (Estimated): September 1, 2024
• Study Completion (Estimated): September 1, 2024

Study Registration Dates

• First Submitted: June 3, 2021
• First Submitted That Met QC Criteria: June 3, 2021
• First Posted (Actual): June 9, 2021

Study Record Updates

• Last Update Posted (Estimated): January 17, 2024
• Last Update Submitted That Met QC Criteria: January 15, 2024
• Last Verified: January 1, 2024

More Information

Terms related to this study

• Keywords: NICU; Neonatal Sleep
• Additional Relevant MeSH Terms: Mental Disorders; Nervous System Diseases; Sleep Wake Disorders; Dyssomnias; Parasomnias; Infant, Newborn, Diseases
• Other Study ID Numbers: HUM00187386; 1R61HL154095-01 (U.S. NIH Grant/Contract)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No