LIVEBORN – mHealth for improved newborn resuscitation

Nearly one million newborns die each year from failure to breathe at birth. Most of these deaths occur in low- and middle-income countries (LMICs). Most are preventable with basic resuscitation consisting of stimulation, suctioning if needed, and bag-mask ventilation (BMV) by one minute after birth. Among these practices, timely and continuous BMV has the greatest impact on mortality. Every 30-second delay in BMV increases the risk of death by 16%; pauses in BMV increase the risk by 75%.

Although resuscitation science and education have reduced newborn mortality, further reductions in mortality require improving local implementation of resuscitation. Since 1992, the International Liaison Committee on Resuscitation (ILCOR) has produced and updated evidence-based guidelines for resuscitation care, systematizing the incorporation of science into newborn resuscitation. In 2010, the American Academy of Pediatrics (AAP) released a simulation-based resuscitation training program based on this science entitled Helping Babies Breathe (HBB). HBB teaches basic resuscitation skills for low-resource settings, and has been widely adopted with approximately one million birth attendants trained in 80 LMICs. A meta-analysis of HBB implementation studies demonstrated a 30% reduction in 1-day newborn mortality. However, despite training, birth attendants struggle to follow HBB practices at the bedside, resulting in on-going gaps in quality resuscitation care and missed opportunities to prevent mortality. For example, in our clinical trial with HBB-trained midwives in the Democratic Republic of the Congo (DRC), birth attendants initiated BMV at a median of 347 seconds (> 5 minutes) after birth. We also noted that stimulation and suctioning contribute to both delayed and interrupted BMV: birth attendants spent a median of 132 seconds stimulating and 98 seconds suctioning during resuscitations involving BMV. 

Innovative tools to support local implementation of evidence-based resuscitation are urgently needed to reduce newborn mortality. Improving local implementation of resuscitation requires a detailed understanding of bedside resuscitation practices. Delivery registers in low-resource settings are insufficiently detailed to understand these bedside practices. For example, in health facilities in the DRC, the information on newborn resuscitation is limited to a dichotomous response of either yes or no. In addition to understanding bedside resuscitation practices, strategies beyond simulation training are required to help birth attendants translate their knowledge and skills to the bedside for public health impact. One promising strategy for improving resuscitation is feedback either during care (i.e., real-time guidance) or after care (i.e., debriefing). In high-income countries, these strategies are typically implemented at the bedside by expert clinicians and augmented by detailed data on provider actions and the newborn’s response during the resuscitation. Lack of expert clinicians in the LMIC setting is a significant barrier to implementing this type of bedside feedback in low-resource settings. The use of mobile health technology enables granular data collection on resuscitation practices and automated delivery of feedback to improve local implementation of newborn resuscitation in LMICs.

LIVEBORN is a mobile health application to support local implementation of evidence-based resuscitation in LMICs through automated feedback. 

We initially developed LIVEBORN as a data collection tool to help health facilities, supervisors, birth attendants and researchers understand what was happening at the bedside during newborn resuscitations. LIVEBORN allowed an observer to collect granular data describing actions of birth attendants and the breathing efforts of the newborn. LIVEBORN also incorporated newborn heart rate data from our new heart rate meter, NEOBEAT, using Bluetooth technology. NEOBEAT is a reusable, battery-operated heart rate meter designed for low-resource settings. NEOBEAT is applied quickly on the newborn abdomen by a single provider at the time of birth. Heart rate from NEOBEAT enables the rapid identification of depressed newborns in need of BMV.

Following data collection for a resuscitation, LIVEBORN displayed a pictorial timeline of the resuscitation events from birth that was reviewed by the observer. When research staff used this tool during a clinical trial in the DRC, birth attendants began asking to view the timeline from LIVEBORN to reflect on their care following a resuscitation. From their interest in data-driven improvement of resuscitation care, the idea for LIVEBORN as a quality improvement tool was born.

In partnership with Congolese midwives, we further developed and evaluated LIVEBORN to support birth attendants with feedback. LIVEBORN integrates data on the actions of the birth attendant and the breathing and heart rate of the newborn to provide real-time guidance during a resuscitation and to support debriefing after a resuscitation.

Real-time guidance

During a resuscitation, LIVEBORN continuously compares the provider’s actions with actions recommended by the HBB algorithm under similar circumstances based on the newborn’s breathing and heart rate. When actions deviate from recommended actions, auditory and visual prompts are provided by LIVEBORN. In partnership with Congolese midwives, we iteratively refined this real-time guidance based on two rounds of simulated resuscitations (Summer 2021). In a feasibility study in one health facility in the DRC, midwives found real-time guidance to be both usable and feasible (Winter-Spring 2022).

Debriefing

After a resuscitation, LIVEBORN produces a pictorial timeline that displays the events of the resuscitation and identifies opportunities for improvement according to the HBB algorithm. Based on formative research with Congolese midwives, the language in LIVEBORN debriefing is carefully crafted to promote improvement rather than blame with the presentation of recommended care, an observation and a discussion point. In partnership with Congolese midwives, we iteratively refined LIVEBORN debriefing based on two rounds of simulated resuscitations (Summer 2021). In a feasibility study in one health facility in the DRC, midwives found debriefing to be both usable and feasible (Winter-Spring 2022).  

LIVEBORN complements the current evidence-based strategy of HBB training paired with low-dose high-frequency simulation practice, supporting birth attendants in consistently translating their knowledge and skills to the bedside. HBB-trained birth attendants in a variety of LMIC settings are currently using LIVEBORN in their clinical care, with a variety of people collecting observational data with the app including supervisors, colleagues, trainees, non-professional facility staff and personnel external to the facility. 

Our target population is newborns who die from failure to breathe at birth, a potentially preventable cause of neonatal mortality that predominantly occurs in LMICs. Approximately 25% of deaths during the first month after birth result from respiratory depression (i.e., failure to breathe at birth). Over 90% of these deaths occur in LMICs. The current trajectory of neonatal mortality globally projects a target of 13.7 neonatal deaths per 1,000 live births by 2030, exceeding the Sustainable Development Goal target of 12/1,000 live births. Given the proportion of newborns who die from respiratory depression each year, targeting this preventable cause of mortality for this underserved population is an urgent public health need. A portion of these deaths result from a lack of health care infrastructure and resources. However, many could be prevented using existing resources by improving the resuscitation care provided by skilled birth attendants.

Our solution equips educated and skilled birth attendants with the support necessary to translate their knowledge and skills into appropriate and timely resuscitation care for vulnerable newborns. Skilled birth attendants in low-resource settings resuscitate newborns under challenging circumstances that include limited training compared to birth attendants in high-income countries, lower staffing to patient ratios, and higher rates of respiratory depression due to poor intrapartum care. As identified by Congolese midwives, these circumstances result in anxiety and stress, ultimately affecting resuscitation performance. While birth attendants in high-resource settings benefit from bedside tools such as cardiorespiratory monitoring and expert coaching, these strategies have been inaccessible to birth attendants in LMICs due to cost. LIVEBORN provides an accessible solution for feedback at the bedside in low-resource settings. Through the use of LIVEBORN we expect birth attendants to more frequently initiate BMV by one minute after birth, and continue BMV without interruption until the newborn breathes. If LIVEBORN feedback reduces the time to BMV by two minutes, prior studies suggest this will lower the risk of neonatal death and hospitalization by 65%.

Laerdal Global Health (LGH) was established to help reach the Sustainable Development Goals for both maternal and newborn health. Since 2010, LGH has worked with the AAP to develop a comprehensive suite of educational programs for birth attendants in LMICs. These programs were developed with consultation and review from international partners around the world. Most notable among these programs is the Helping Babies Breathe program, which has been translated into 25 languages and has been used as the training program for birth attendants and other pediatric providers in 80 countries. It is estimated that more than 1,000,000 providers have completed this training. LGH’s commitment to innovation and experience developing and marketing tools for skilled birth attendants in low-resource settings make LGH well-positioned to deliver LIVEBORN.

In addition to technology experts from LGH, our team for LIVEBORN includes neonatologists from the University of North Carolina at Chapel Hill and public health practitioners from the Kinshasa School of Public Health. Following a history of working together on curriculum development and clinical research, in 2017 our collaborative team established an interest in implementation science and began to develop a “learning laboratory” for newborn resuscitation in three health facilities in Kinshasa, DRC. Through the evaluation of our battery-operated heart rate meter, NEOBEAT, in these facilities, we developed a strong partnership with the Congolese midwives staffing these facilities and have welcomed them to our research and development partnership.

In partnership with Congolese midwives in our learning laboratory, we developed and evaluated LIVEBORN. Specifically, we began our journey by conducting focus group discussions to understand challenges to implementation of basic resuscitation, preferences for real-time guidance and debriefing, and facilitators and barriers to using LIVEBORN in clinical practice. Based on this formative work with 24 Congolese midwives, we developed LIVEBORN. We tested LIVEBORN in simulated resuscitations with these same midwives, and iteratively refined its real-time guidance and debriefing based on their feedback and experience, producing a product that was usable by their assessment.

Recognizing the potential staffing challenge for observing resuscitations with LIVEBORN in these facilities, Congolese midwives led the strategy development for incorporating LIVEBORN into their facilities. Using a methodology called Trials of Improved Practices, the midwives refined their strategy until they developed a feasible solution to incorporate LIVEBORN into their clinical practice. For example, at one facility, midwives conduct observations and utilize LIVEBORN debriefing both to consider the quality of their own care and to assist nurse trainees to reflect upon how they can improve. In a second facility, in addition to midwives, environmental health services staff conduct observations with LIVEBORN to ensure that real-time guidance is available for the majority of resuscitations. 

As a part of this challenge, we want to make LIVEBORN ready for broader consumer use by

• Collecting data and exploring technical solutions (including AI) for automatic registrations, to make LIVEBORN less dependent on the observer and thereby ease scale-up
• Making the system less reliant on back-end support, to reduce implementation cost
• Conducting implementation research alongside implementation in new sites, to get a better understanding of learning cultures and barriers to debriefing and real-time guidance
• Running business model innovation workshops, to inform the business plan for further scale up 

Since the publication of the Millennium Development Goals in 2000, reinforced by the Sustainable Development Goals in 2014, international health organizations, most notably the World Health Organization, acknowledged the critical need to improve newborn survival. Since the endorsement of these goals, the primary strategy for reducing mortality has been to educate birth attendants and newborn care providers about evidence-based care including resuscitation. However, we may be approaching the maximum impact of education alone. Further reductions in mortality will require new and innovative strategies that facilitate the translation of education into practice.

LIVEBORN offers a new and innovative approach to improving the translation of resuscitation knowledge and skills into practice. LIVEBORN provides the platform for collecting granular data describing the physiology of the newborn and actions of the provider. Our user-centered design of the event registration screen allows data to be collected by a variety of personnel, including professionals and non-professionals. Additionally, LIVEBORN automatically identifies gaps in the quality of resuscitation care. It delivers feedback to providers, both in real-time and immediately after the event. This feedback is automated, reducing the need for clinical experts at the bedside. These attributes make the use of LIVEBORN compatible with virtually all facility-based birth environments in LMICs.

We expect LIVEBORN to drive the improvement of newborn resuscitation beyond training with simulation practice to a new focus on bedside learning. The granular data collected by LIVEBORN will support innovations in data-driven training, auditing, and continuous quality improvement for improved performance of newborn resuscitation. The real-time guidance delivered by LIVEBORN will prompt additional innovations in clinical decision support during resuscitation. LIVEBORN debriefing will facilitate post-event reflection, contributing to a facility culture of learning rather than one of blame.

Our ultimate impact goal is to reduce newborn mortality from respiratory depression by making every newborn resuscitation a learning event.

Our solution is in a growth phase, as demonstrated by its use in 19 facilities representing five countries (the DRC, Ethiopia, Nepal, Norway, Tanzania).  Our overall impact goal for the next year is to double the reach of LIVEBORN in these same countries. In concert with this expansion, we have planned several evaluations of the effectiveness of LIVEBORN feedback, as this functionality is currently only used as part of research projects. During the next year we will systematically expand the use of LIVEBORN to a consortium of eight hospitals in Kinshasa that will support an evaluation of the effectiveness of the use of LIVEBORN in a randomized, controlled trial (Funding: Phased Innovation Award from NICHD, R33 funding anticipated). We will also evaluate the effectiveness of LIVEBORN feedback in pre-post studies in a large health facility in Tanzania and a large hospital in Nepal (Funding: World Bank Global Financing Facility Innovation to Scale award; Swedish Research Council [pending]). This expansion of LIVEBORN coupled with evaluations of its effectiveness will demonstrate its adaptability to a variety of settings in LMICs, preparing a foundation of experience for scale-up.

Our overall impact goal for the next five years is to scale-up the use of LIVEBORN on a national level in several LMICs. Our current partnerships in the DRC and Tanzania linked to their respective ministries of health will facilitate national scale-up in these settings. Key to our progress towards scale-up, we will incorporate evaluations of the feasibility, acceptability and sustainability of LIVEBORN into the studies we initiate on effectiveness in the upcoming year. Multi-site data on these implementation aspects will allow us to adjust LIVEBORN to ensure successful scale-up. 

We are measuring our progress towards our impact goals with the following indicators: 

• Number of facilities using LIVEBORN annually: We want to expand the impact of LIVEBORN by reaching more birth attendants. As we expect LIVEBORN to be implemented at the facility level, we are actively charting the number of facilities using LIVEBORN annually.

• Number of births registered with LIVEBORN per month: Our ultimate impact goal is to reduce newborn mortality, thus we want to expand the reach of LIVEBORN to more births. The LIVEBORN cloud allows us to monitor the number of births registered with LIVEBORN each month.

• As we begin formal evaluations of the effectiveness of LIVEBORN later this year, we will begin charting additional indicators related to birth attendant performance such as the time to bag-mask ventilation and newborn outcomes including neonatal mortality.

From our ‘learning laboratory’ with Congolese midwives and the use of LIVEBORN to collect granular data on resuscitation practices, we understand that trained birth attendants underutilize BMV as a strategy to help babies breathe. When they do initiate BMV, it is significantly delayed due to other practices such as stimulation and aggressive suctioning. From our focus group discussions with Congolese midwives on challenges to implementing basic resuscitation, we understand that the stress of resuscitating a flaccid infant affects birth attendants’ ability to access their knowledge and skills to provide quality care at the bedside.

LIVEBORN will provide feedback both during and after resuscitation, helping birth attendants to engage in bedside learning to improve their care. Our development of LIVEBORN, specifically its use of bedside feedback to improve BMV, was informed by the Cognitive Load Theory. Cognitive load is the total amount of mental effort used in working memory. The Cognitive Load Theory states that cognitive load is made up of intrinsic, germane, and extraneous components. Novice providers performing BMV experience high cognitive load due to the inherent difficulty of BMV (intrinsic load) combined with the stress of resuscitating a non-breathing newborn (extraneous load). This situation results in default to procedures that present less cognitive load (e.g., suctioning) even if these procedures are ineffective. Excessive cognitive load leaves limited room for learning (germane load). We developed LIVEBORN to capitalize on two strategies to ensure efficient learning in situations requiring high cognitive load: decreasing extraneous load and optimizing germane load. LIVEBORN decreases extraneous load by using real-time guidance to reduce stress associated with resuscitating a depressed newborn. LIVEBORN optimizes germane load by using debriefing to help skilled birth attendants learn from their clinical care.

We expect LIVEBORN feedback to both increase the utilization of BMV and reduce the time to BMV, with the additional benefit of reducing aggressive suctioning. Feedback during and after cardiopulmonary resuscitation is a commonly used tool to improve clinical care in high-resource settings. While feedback strategies are under-studied, the literature suggests that cardiopulmonary resuscitation performance is improved using these strategies. Based on clinical experience and limited evidence in the literature, we expect birth attendants to improve their care in response to feedback. As the feedback delivered by LIVEBORN targets appropriate and timely use of BMV as well as reduced suctioning, we expect these metrics to improve with the use of LIVEBORN.

We expect LIVEBORN feedback to reduce newborn mortality from respiratory depression. There is strong evidence of the impact of appropriate and timely BMV on newborn survival at birth in low-resource settings. For example, if LIVEBORN feedback reduces the time to BMV by two minutes, prior studies suggest this will lower the risk of neonatal death and hospitalization by 65%.

The core technology powering LIVEBORN is software, mobile applications and an innovative medical device. Development and implementation is driven by evidence-based knowledge and service design to facilitate easy data collection and improve newborn resuscitation.

The Liveborn App is used for data collection, real-time guidance (audio) and debrief. The app can connect to NeoBeat and receive heart rate data when NeoBeat is applied to a newborn. Real-time guidance can be enhanced to audio-visual feedback by using the Guidance app. In addition, data is uploaded to a secured cloud and can be accessed and shared in a web solution. The web solution includes visual dashboards for tracking key numbers over time for trending and quality improvement as well as settings for data sharing/access.

The Liveborn App allows data collection during newborn resuscitation, gives real-time guidance and facilitates debrief. Guidance and debrief are rule-based, directly implemented in the app and available offline. When internet connectivity is available, data is transferred to the secured cloud storage and features on the web interface (access raw data, sharing data, viewing QI dashboards) can be used.

Data transfer and access is secured, user access based on authentication and authorization. GDPR principles are implemented by design and authorization is granted only after data processing agreement is signed.

The rules for audio-visual real-time guidance and debrief, as well as their presentation to the user are designed to trigger favored behavioral change.

The clinical data that informs LIVEBORN feedback is collected by an observer in real-time starting with the time of birth. During our early use of LIVEBORN as a research data collection tool, we tasked paid research nurses with the responsibility for collecting data. When we enhanced and repurposed LIVEBORN as a clinical, quality improvement tool, we were challenged with assigning the responsibility for observation and data collection to personnel available in a clinical service. We began by simplifying the registration screen and developing a training program for observes that included practice registering events with a set of newborn resuscitation videos. Through our formative work with Congolese midwives, we learned that this training program sufficiently prepares professionals to accurately observe resuscitations using LIVEBORN. While these midwives recognized the value of incorporating LIVEBORN into their clinical care, the low clinical staffing resulted in times when a midwife was not available to observe the care of a colleague. As such, midwives emphasized the importance of off-loading observations from health facility staff in order to enhance scale-up of the LIVEBORN system.

During our formative work, Trials of Improved Practices, we challenged the midwives to consider creative solutions to this problem. The midwives developed and tested innovative strategies that allowed them to feasibly incorporate LIVEBORN into their clinical practice. For example, at one facility, midwives worked in pairs to evaluate their performance when the burden of clinical service permitted, but also observed resuscitations led by nurse trainees when the service was busy. This strategy had the advantage of permitting more opportunities for midwives to understand the quality of their care but also assisted trainees in reflecting upon how they could improve. In a second facility, in addition to midwives, environmental health services staff conducted observations with LIVEBORN to ensure that real-time guidance was available for the majority of resuscitations. It is our plan to use these and potentially other creative strategies for the observation of deliveries as we expand the use of LIVEBORN in our growth phase.

In acknowledgement of the limitations of the requirement for an observer, we are actively exploring possibilities to reduce the dependency of the observer including automatic interpretation of video footage of resuscitations using artificial intelligence, automatic detection of breathing and automatic detection of positive pressure ventilation using sound.

Diversity, Equity and Inclusion are important values to Laerdal Global Health.We seek to have a gender-balanced team, although currently we have a team where only 1/3rd are men. We are therefore actively seeking to employ more men in future positions, in particular in the teams where women are dominant. Our board is consisting of 50% men/women. We have a female CEO, a management team with 75%women. In our team we have 13 nationalities represented from four different continents, this is an important asset to the team and enables us to do our jobs better. We actively seek this diversity in recruiting. It is also important that our company is a place to grow and that all of our staff gets the chance to develop professionally. Including taking on new challenges and positions within the team. We have professional development ladders where you can advance both in taking on managing roles and also senior positions as technical experts. Our managers have all started in other roles and been promoted as part of their professional development.

We offer financial support to continued professional development and encourage team members to explore and enroll in formal education, courses and training that can help them develop.

Diversity, equity and inclusion are also important values guiding us in our development process.

https://laerdalglobalhealth.com/about-us/design-process/ , we focus on human centered design and work in an iterative way where we seek to involve all types of users and stakeholders.

An example is how we strive to also understand the perspectives of the mothers and families, when developing a new solution for health workers. 

Laerdal Global Health sells solutions to implementors of health programs in order to help them become more efficient and get increased impact of their programs. Our main customers are UN organizations, international and national NGOs, research groups and in an increasingly national and local governments.  

We have a portfolio of educational programs, skills trainers and simulators, some key medical devices  and increasingly mHealth applications for data collection and quality improvement. In addition, in countries where we have local presence, we provide services related to implementation and training.

Most of our products are sold through our customer service team, or through distributors in selected countries we also have long term agreement with Unicef and UNFPA, and other INGOs for much of our portfolio. At the moment, digital products are bundled with physical products, but we are moving into providing software as a service (SaaS) and subscription business models for our solutions.

For LIVEBORN we wish to offer this as part of other established product bundles such as Helping Babies Breathe (HBB)  bundle https://laerdalglobalhealth.com/partnerships-and-programs/helping-babies-breathe/  and Safer Births Bundle of Care https://saferbirths.com/innovations/. Building on the customer base that we already have for these programs. We know that the current implementors of these bundles are seeking solutions to help transform increased competence from training into change in behavior in clinical practice and ways to collect data for use in quality improvement activities.

This will increase the impact of their programs even further. 

For 10 years Laerdal Global Health has been developing face-to-face training programs, simulators and skills trainers and medical devices for helping save lives of mothers and babies in the low resource setting. These products are sold at a non-for-profit basis to the countries with the highest maternal and newborn mortality.  The revenue model has partially been based on recovering development cost through offering the same solutions at higher prices in high income setting and covering development costs through grants.  We also believe that our physical products bring more value when offered in a bundle with internationally recognized training programs such as the Helping Mothers and Babies Survive programs, and we see that these programs drive the sales of physical products such as medical devices and simulators for training.

We are now moving into a phase where we seek to maximize the impact of our solutions by complementing them with digital tools for capturing, analyzing, and presenting data to the customers and the users, essentially selling the solutions as a service through subscription rather than standalone products that are one-time purchases.  

LIVEBORN is one of these service solutions that we seek to offer customers, instead of a one-time income of the sales of medical devices or simulators, LIVEBORN will be sold as a subscription to complement these solutions and continue to offer value to the customer over time. Continuously developing new features and making improvements to our mHealth apps, we foresee an increase in value to the customer over time and increased income from the different subscriptions offered. 

In Laerdal Global Health we have seen multiple examples that developing and offering education programs linked to physical devices and training tools helps drive the sales of the physical products, because this increases the value to the customer of the physical products. We therefore believe more in offering complete bundles that to an increasing degree includes services, data packages and tools for quality improvement.

As an example, we developed the Helping Babies Breathe program together with the American Academy of Pediatrics in 2009, in addition to the training curriculum we also developed the NeoNatalie simulator, the penguin suction device and improved versions of the bag and mask. We worked closely with partners in implementation, and supported implementation research and engaged in large alliances (the Survive and Thrive alliance) to drive the spread of the program. The results are that more than 1million birth attendants have been trained, and 163 000 NeoNatalies, 782 620 Penguins and 323 017 resuscitators sold.