Bedside Improvement of Resuscitation Through mHealth Feedback (BIRTH Study)

Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), Doris Duke Charitable Foundation, Laerdal Global Health, Kinshasa School of Public Health

Purpose: To evaluate the effectiveness of newborn resuscitation feedback supported by a mobile health application called LIVEBORN; secondarily, to evaluate the relative effectiveness of real-time guidance vs debriefing. Participants: Newborns and Nurse midwives Procedures (methods): This is a pre-post interventional trial to evaluate the effectiveness of LIVEBORN feedback. The investigators will use a randomized design to test the relative effectiveness of two modes of feedback: real-time guidance versus debriefing. Given the potential for feedback interventions to have spillover effects, the investigators will randomize by cluster (i.e., facility) rather than by individual. The study will begin with an approximately two-month pilot phase to establish systems for implementation of recommended training and simulation practice and consistent use of LIVEBORN for observations. After these systems have been successfully established, the investigators will initiate the pre-post trial. The control phase will last six months followed by implementation of LIVEBORN feedback in an intervention phase lasting 12 months.

Study Sites: the investigators will conduct this pre-post interventional trial in six health facilities in Kinshasa, Democratic Republic of the Congo (DRC). Pilot Phase: The purpose of the pilot phase is to establish systems to ensure 1) implementation of recommended training and simulation practice and 2) consistent use of LIVEBORN for observations. The pilot phase will last approximately two months. Any facilities deemed facile with LIVEBORN will be exempted from the pilot phase. Establishment of a System for Implementation of Recommended Training and Simulation Practice: During the pilot phase, the investigators will work with each facility to ensure a system for implementation of recommended resuscitation training and low-dose high-frequency (LDHF) practice of bag-mask ventilation (BMV). Basic Resuscitation Training: The investigators will train all nurse midwives in basic resuscitation using Helping Babies Breathe (HBB) 2.0 materials adapted to include the use of NeoBeat, a battery-operated heart rate (HR) meter. In addition to reviewing the evidence-based HBB resuscitation algorithm, research staff will orient nurse midwives to use of NeoBeat. The training will also include how to use HR to accurately distinguish liveborn from stillborn infants. The investigators will recommend that nurse midwives use NeoBeat for high-risk pregnancies and newborns not crying by 30 seconds. Establishment of a System for Consistent Use of LIVEBORN for Observations: During the pilot phase, the investigators will work with each facility to ensure a system for consistent use of LIVEBORN for observations. This system will include placement of NeoBeat on newborns requiring resuscitation. The investigators will use an established participatory research methodology called trials of improved practices (TIPs) which the investigators successfully used in the R21 phase of this work to design a strategy for consistent use of LIVEBORN for observations. At each facility, the investigators will conduct a strategy development session with facility leadership such as the head nurse midwife to develop an initial strategy for implementation of LIVEBORN. Each health facility will implement the locally-developed initial strategy in a small-scale test for approximately three weeks. Throughout the small-scale test, the investigators will rapidly analyze data and identify barriers to implementation of the strategy as well as potential solutions. The investigators will refine the strategy in additional cycles of TIPs (strategy development, small-scale testing, rapid analysis) until the investigators have identified a feasible strategy to move into the control phase. Control Phase: The purpose of the control phase is to gather detailed, prospective data on resuscitation care and newborn outcomes while implementing recommended practice. During a six-month control phase, midwives will implement HBB in their clinical practice with LDHF simulation practice of BMV. LDHF practice will occur at regular intervals per the system set up during the pilot phase. Nurse midwives will place NeoBeat on newborns that they anticipate resuscitating. Data Collection: Research staff will conduct medical record abstraction to document clinical data for each eligible newborn including such elements as maternal parity and age, birth date and time, infant gestational age, birth weight, mode of delivery, stillborn or liveborn, alive or dead at discharge (including date of death). As in the pilot phase, trained observers (research staff, midwives, and other designated observers per the strategy developed in the pilot phase) will conduct observations of resuscitations for a convenience sample of births using the LIVEBORN app. If NeoBeat is placed on the newborn, the investigators will collect ECG data. Preparatory Activities for the Intervention Phase: While in the control phase, the investigators will randomize facilities to either real-time guidance or debriefing with LIVEBORN in preparation for the intervention phase. Prior to implementation of the intervention and towards the end of the control phase, research staff will prepare nurse midwives to implement LIVEBORN feedback per their cluster randomized assignment using simulation. Intervention Phase: The purpose of the intervention phase is to gather prospective data on resuscitation care and newborn outcomes during implementation of LIVEBORN feedback. During this 12-month phase, nurse midwives will implement LIVEBORN feedback per their cluster-randomized assignment to either real-time guidance or debriefing. Additionally, nurse midwives will continue all study procedures from the control phase, including implementation of HBB with LDHF practice, use of NeoBeat on newborns requiring resuscitation, and observation of births using LIVEBORN. Implementation of LIVEBORN Feedback: At facilities randomized to real-time guidance, midwives will be supported by audio-visual feedback from LIVEBORN during the convenience sample of resuscitations observed with LIVEBORN. At facilities randomized to debriefing, midwives will debrief with LIVEBORN for at least 50% of the convenience sample of resuscitations observed with LIVEBORN that involved a newborn not crying by 60 seconds after birth and/or the receipt of BMV. Data Collection: As in the control phase, research staff will conduct medical record abstraction to document clinical data for each enrolled. As in the pilot and control phases, trained observers (research staff, midwives, and other designated observers per the strategy developed in the pilot phase) will conduct observations of resuscitations for a convenience sample of births using the LIVEBORN app. If NeoBeat is placed on the newborn, the investigators will collect ECG data. At the end of the intervention, five midwives & the head nurse midwife from each facility will complete surveys consisting of 4-item quantitative scales to evaluate feasibility and acceptability using the Feasibility of Intervention Measure (FIM) and Acceptability of Intervention Measure (AIM).

The LIVEBORN app will provide audio-visual guidance to the birth attendant during a resuscitation based on the data inputted by an observer as well as heart rate data from NeoBeat (if used).

The LIVEBORN app will support birth attendants in debriefing following a resuscitation based on the data inputted by an observer as well as heart rate data from NeoBeat (if used).

LIVEBORN is a mobile health application designed to help providers learn from their clinical practice during newborn resuscitation through real-time guidance and debriefing. Using LIVEBORN, an observer can document events during a resuscitation such as the actions of the provider and when the baby cries. A battery-operated heart rate meter, NeoBeat, measures and sends the newborn's heart rate to the LIVEBORN app via Bluetooth. LIVEBORN integrates data from the observer and NeoBeat to provide audio-visual real-time guidance to the provider during a resuscitation. Following the resuscitation, providers can also debrief using LIVEBORN. The app operates on the Android system and is functional off-line, except for the need for an internet connection to upload data to the cloud for research.

NeoBeat is a low-cost device for the measurement of newborn heart rate (HR) that was developed by Laerdal Global Health for use in low-resource settings. It is applied by placing the device around the torso of the newborn. It detects HR accurately, quickly (<5 sec), is reusable and can be easily disinfected. NeoBeat's dry electrodes detect HR which is digitally displayed and recorded. It uses an impedance system to sense skin contact and motion detection to validate signal quality. It does not require the use of buttons, adhesive, gel or cables and has no disposable parts. A single birth attendant can apply NeoBeat in 2-3 seconds.

The investigators will measure duration of suctioning as a continuous variable among liveborn neonates who are not crying by 30 seconds. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure duration of suctioning as a continuous variable among liveborn neonates who are not crying by 30 seconds. This variable will be collected via direct observation using the Liveborn application.

The investigators will report the proportion of livebirths not crying by 30 seconds who are suctioned for less than 30 seconds. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure time to BMV as a continuous variable, defining time to BMV as the number of seconds between birth and the start of BMV. They will report time to BMV among all livebirths not crying by 30 seconds who receive BMV. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure time to BMV as a continuous variable, defining time to BMV as the number of seconds between birth and the start of BMV. They will report time to BMV among all livebirths not crying by 30 seconds who receive BMV. This variable will be collected via direct observation using the Liveborn application.

Proportion of Newborns Who Receive BMV within the Golden Minute Among All Livebirths Not Crying by 30 Seconds Who Receive BMV

In addition to time to BMV, a dichotomous variable of BMV initiation within 60 seconds after birth will be measured. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure a dichotomous variable of receipt of BMV at any point during delivery room resuscitation among liveborn infants who do not cry by 60 seconds after birth. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure the time to HR≥100bpm as a continuous variable among livebirths who have not cried by 30 seconds after birth who have NeoBeat placed. This variable will be collected using NeoBeat.

The investigators will measure the time to HR≥100bpm as a continuous variable among livebirths who have not cried by 30 seconds after birth who have NeoBeat placed. This variable will be collected using NeoBeat.

The investigators will measure time from birth to first cry as a continuous variable among liveborn infants who have not cried by 30 seconds after birth. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure time from birth to first cry as a continuous variable among liveborn infants who have not cried by 30 seconds after birth. This variable will be collected via direct observation using the Liveborn application.

The investigators will also look at the total duration of suctioning among all livebirths who cry by 30 seconds after birth. This variable will be collected via direct observation using the Liveborn application.

The investigators will also look at the total duration of suctioning among all livebirths who cry by 30 seconds after birth. This variable will be collected via direct observation using the Liveborn application.

The investigators will also look at a dichotomous variable of the proportion of newborns who receive less than 30 seconds of total suctioning among liveborn infants who cry by 30 seconds after birth. This variable will be collected via direct observation using the Liveborn application.

The investigators will measure 24-hour newborn mortality as a dichotomous variable for all enrolled newborns. In the participating health facilities, typical length of stay following birth is more than 24 hours. In the rare instance when a newborn is discharged home prior to 24 hours after birth, the investigators will assume the newborn is alive at 24 hours. If the newborn is transferred prior to discharge, the investigators will contact the referral facility to determine 24-hour vital status.

The investigators will define fresh stillbirths based on the provider's documentation in the medical record. The investigators will measure perinatal mortality as a dichotomous variable for all enrolled newborns.

The investigators will review the subset of stillbirth cases with heart rate (HR) data from NeoBeat to determine whether the case is stillborn or liveborn based on HR. The investigators will define misclassification as cases identified by a provider as stillborn who have a HR per NeoBeat. The investigators will measure the dichotomous outcome of stillbirth misclassification.

Newborns admitted to a participating health facility who meet the following criteria: Inclusion criteria: All in-born neonates, either liveborn or intrapartum (i.e., fresh) stillborn, regardless of multiple gestation or maternal complication. Exclusion criteria: Less than 28 weeks completed gestation (or if gestational age is unknown, birth weight <1,000 grams), outborn, known congenital anomaly, antepartum (i.e., macerated) stillbirth. Nurse midwives employed at a participating health facility who meet the following criteria: Inclusion criteria: All midwives and nurses functioning in the role of a midwife who care for newborns at the time of birth during the course of their regular employment. Exclusion criteria: Unwillingness to consent

Data will be made indefinitely available no later than the acceptance for publication of the main findings from the final dataset.