How to Improve the Clinical Delivery of Neonatal Resuscitation at Birth with a Human-Centred Design Approach
Type: Master thesis
Timeline: 6 months(Jan. — Jun.2021)
Contribution: Interview, Persona, Journey Map, HMW, Brainstorming, Storyboard, Testing
Supervisor: Giovanni Pignoni
Co-supervisor: Michael Site(Laerdal Medical), Siren Irene Rettedal (SUS)
01 Research Problem
Each year around 136 million babies were born, while most babies can breathe independently within the 30s after birth, around 10% of these babies need some assistance to establish a successful respiratory transition, such as stimulation, ventilation and very rarely chest compression.
However, studies show a high percentage of non-adherence to the guidelines and high error rates when healthcare professionals (HCPs) resuscitate the newborns, which can lead to sub-optimal outcomes and even morbidity and mortality.
02 Research Questions
Therefore I chose this as my thesis topic and here are some research questions that I wanted to find out.
- What are the barriers and enablers for providing a timely, safe and effective newborn resuscitation?
- What kind of needs and expectations do healthcare professionals have?
- How can we support healthcare professionals to improve the quality of care?
If there’s a solution for it, how can the solution be proven to be effective? And what is the requirement for implementation?
03 Discover
Based on my educational background and personal interest, a human-centred design approach has been chosen and can be visualised in a double-diamond model.
In the discover stage, I started with a literature review on how HCPs perform resuscitation, what guidelines they follow, what difficulties they have, and what factors can contribute to a good outcome.
3.1 Literature Review
I have studied several guidelines in different countries. This is the diagram that I made based on the 2015 Norwegian resuscitation guideline.
Before birth, the team should have a brief to check the equipment and assign tasks. After the baby is born, if it’s not a full-term baby, not crying or breathing or has a bad tone, then it should be dried and stimulated, the heart rate (HR) and breathing should be checked. If HR is below 100beats per minute, the team should start the ventilation for 60 seconds. If the HR is below 60 bpm, they should start chest compression (CC) together with ventilation.
In this time-critical situation, many errors could happen. For example:
- Assess HR inaccurately through auscultation and palpation;
- Incorrect ventilation pressures;
- Interruption of ventilation to stimulate or check the HR;
- Clinically significant delays in initiating ventilation
Research found that assessing HR and ventilation techniques based on human judgements are inaccurate and unreliable. and if they get the wrong HR, the following procedures based on HR could go wrong too.
Other reasons that cause mistakes are lack of knowledge and skills, high cognitive load and technical load. The stress and sense of time pressure make it difficult for HCPs to remember the steps in the guidelines, skip certain steps or proceed to the next step without indication.
Enablers for High Quality Neonatal Resuscitation
Luckily, there are some factors that can make resuscitation easier.
- Decision support tools (DSTs)
DST is a tool that often includes visual displays and auditory prompts to remind the HCPs to take action, these reminders can reduce part of the cognitive load and save more capacity for other tasks.
- Respiratory Function Monitors (RFMs)
Respiratory function monitors (RFMs) can measure and monitor many parameters such as HR, oxygen saturation, airway pressure, gas leakage and so on, allowing the resuscitator to discover problems and adjust their technique.
- Briefing and Debriefing
Studies show that briefing, debriefing and checklist can improve team communication, facilitate teamwork, and can improve resuscitation outcomes.
What is briefing:
- Determine the leader and assign the roles and tasks;
- Check and prepare the equipment;
- Discuss the treatment plan with the parents if not already done;
What is debriefing:
- Debriefing normally happens after a simulated or clinical event.
- The goal of debriefing is to improve future performance.
3.2 Field Study
When I conducted the interviews at Gjovik hospital and Stavanger University Hospital (SUS), I had the chance to visit the resuscitation rooms. These two pictures show what equipment was in the resuscitation rooms and their locations.
3.3 Online Survey
Since it’s challenging to reach the doctors and nurses in different hospitals, I have conducted an online survey which can easily reach a large group of people, and doesn’t require too much effort for the participants to complete a survey. I have also used the survey as a way of recruiting participants for interviews, and to adjust my interview questions based on the survey results.
I have received 47 valid submissions from doctors and nurses in different hospitals. most of the participants came from SUS and St. Olavs Hospital in Trondheim. Considering the survey was mainly distributed through SUS and St. Olavs Hospital in Trondheim, it’s no surprise that we got these results.
- The largest group of the participants have work experience of more than 12 years.
- 3 of the most collected data are HR, SpO2 and Apgar score.
- The frequency of participation in clinical events and training is quite low.
3.4 Individual Interview
11 doctors and nurses in total participated in a semi-structured interview, most of them were from SUS.
The interview recordings were transcribed into text files and then put into Nvivo for analysis. Before the coding process, some categories had been created based on the interview questions, such as “how they perform newborn resuscitation” and “what difficulties they have”. More categories were developed during the coding process. The interview notes were divided into small pieces of meaning units, abstracted and labelled with a code. Similar or related codes were grouped under the same subcategory, and several subcategories were then grouped under a bigger category, as shown in the figure below. This coding process helped me to retrieve information across different interview files to later create persona, empathy map and journey map.
Comparing Results from Survey and Interview
Results from the online survey showed that 86% of the HCPs who had clinical experience had difficulties during the clinical event.
The four of the most chosen difficulties in the survey are shown in the figure below . However, when I further investigated it in interviews, the most chosen difficulties were “the difficulty to get sufficient air into the lungs” and “communication difficulty”. These two difficulties were chosen to follow up in the later stage.
04 Define
In the define stage, persona, empathy map, and journey map have been used to visualise the insights. 2 main difficulties have been identified and written in Point of View and How Might We, which have been used in the co-creation workshop in the develop stage.
4.1 Persona & Empathy Map
Findings from the survey and interview were presented as persona, empathy map and journey map. I have created five personas for pediatrician, NICU nurse, midwife, anesthesiologist and anesthesia nurse. Each persona has a corresponding empathy map. The persona and empathy map helps me to create empathy with the users.
4.2 Scenario and Journey Map
I have created four journey maps based on four main scenarios, which can be categorised as expected situations and unexpected situations.
- Expected situations (happen 2/3 of the time): expected case in the delivery ward, emergency c-section in the operation theatre
- Unexpected situations (happen 1/3 of the time): unexpected case in the delivery ward and planned c-section in the operation theatre
The journey maps helped me understand how different roles work in a team in different scenarios.
05 Develop
5.1 HMW
Two main difficulties, “difficulty of providing sufficient ventilation” and “communication difficulty within the team”, have been written as Point of View and How Might We questions. Here are the two How Might We questions, which served as the basis for the co-creation workshop.
HMW1: How might we [make it easier for the HCPs who participate in a newborn resuscitation to know if they manage to give sufficient ventilation and possibly give them guidance on the next step]?
HMW2: How might we [provide an overview of the situation and help the HCPs find their tasks]?
5.2 Co-creation workshop: Brainstorming and Dot Voting
I have invited one neonatologist and one anesthesiologist to a co-creation workshop to validate the 4 journey maps, to generate ideas for the How Might We questions and then to vote out some ideas.
I also presented three competitors to the participants in the workshop to ask for their feedback. Generally, they gave positive feedback. However, they expressed their concerns about the overload of information. After the workshop, I grouped similar ideas together and developed them into one solution.
06 Deliver
6.1 The Ecosystem of the Resuscitation Support Tool (iPad)
The picture below demonstrates how the resuscitation support tool (run on iPad) interacts with other objects in the ecosystem. The resuscitation support tool receives real-time data from NeoBeat, Monivent Neo 100 and a pulse oximeter. NeoBeat can collect reliable HR within 5 seconds and are currently in use in SUS. Monivent Neo 100 has a sensor module and a digital screen. Its sensor module can be added to the bag mask ventilator and T-piece ventilators, and the digital screen display ventilation parameters such as Vte, PIP, PEEP, mask leakage and ventilation rate. A pulse oximeter is integrated into the Panda warmer and can measure SpO2.
In the future, visual recognition technology can be added to this solution so the time of birth and different intervention measures can be automatically registered.
The resuscitation support tool and Monivent Neo 100 should be connected to a charger and could be mounted on the wall next to the panda warmer with screen brackets, which allow HCPs to adjust the positions and angles for both screens. And the resuscitation support tool could be easily taken down from the screen bracket when in use.
6.2 Storyboard
A storyboard was created based on the most common scenario, where a baby is expected to have a breathing problem before birth, and the doctors and nurses will prepare and wait in the resuscitation room. The storyboard visualizes how the resuscitation support tool can be implemented in their current workflow and has a positive impact. The storyboard together with the prototype has been tested with 2 HCPs in the individual interviews. Due to the limited space, here I only show parts of the whole storyboard.
6.3 Prototype
I made a medium-fidelity prototype for the resuscitation support tool. The prototype has four main features, namely Brief, Resuscitation, History and Setting.
- Brief
On the brief page, you can see a checklist which serves as a reminder that all these tasks should be finished during a brief. The checklist has four items and they follow a natural order, where they introduce themselves first and assign roles, and then discuss the patient history and treatment plan, and then the team leader can express his/her expectation for closed-loop communication, for example asking the teammates to call back the orders (eg. “HR is 90”), and ask the teammates to speak up if they have any problems or concerns. And the last one is to check the equipment and make sure they have the right settings. When you click on the equipment checklist it opens another page with an equipment checklist based on different roles.
During the testing where the prototype and storyboard were tested with 2 HCPs, one participant recommended moving the equipment checklist to the top, giving the reason that they could at least get the equipment ready in case the baby was born earlier than expected.
- Record resuscitation page
when you click on the “baby born” button, the resuscitation page is activated, and the clock starts running. with the segmented control button, you can switch between record mode and overview mode. in the recording mode, you can register the baby’s breathing status, different interventions, apgar score and some birth information. There will be some notifications at the bottom of the page, for example, at one minute there will be a reminder to fill in the apgar score. the participant mentioned that this can increase their awareness of certain procedures.
however, they don’t like that it requires an extra person to record interventions on the iPad and the need of type in a weight, because they don’t weigh the baby until it’s stable.
it might not be able to register the correct time of birth and cord clamp if the iPad is not in the delivery room
in the overview mode, you can see all kinds of data that can aid the decision making process, such as HR, SpO2, End Tidal Volume (Vte), flow, ECO2 and so on. You can also see what interventions have been taken at what time for how long.
After you finish the observation of resuscitation, you can click at the button “end resuscitation” at the top right corner and save it. then you can review it in the history page or go to debrief immediately.
On the debrief page, you can see the summary of the resuscitation, suggestion on what you can improve in the future
and discussion topics. Underneath are some measures that have been taken, and some important data with timelines.
The participants thought that this solution make it easier to assess the baby and can improve the ventilation.
the objective feedback can inform them what has happened during the resuscitation. they especially like the debrief page and said it could make debrief much easier. and there’s a potential of integrating this data into the patient’s journal and using this data to tailor simulation training based on their weaknesses.
- Different versions of record resuscitation page
I have made two version of the record page, on version a, you can see all the possible intervention measures, while on version b, they are separated into two groups, basic interventions and more advanced interventions.
both participants said that they liked the simplicity of version b.
because very few babies need ventilation and more advanced interventions, so there’s no need to show them right away.
the consideration is to reduce the unnecessary contents on the screen in order to reduce the workload for the user.
- Different versions of overview page
I have also made three versions of the overview page. On version a, ECO2 is shown in a wave form and refresh every few seconds, while on version b the ECO2 was shown with the changes in a time line.
both participants said they preferred version a, which has waveform of ECO2. They would also like to see HR and SpO2 in waveform. because waveform is more reliable, but the timelines would be useful in the debrief report.
While both participants shared the same preference for the waveform data, they
disagreed with the background colors of the graphs. While the pediatrician preferred
the version with warning and danger areas highlighted, the anesthesiologist
preferred the version c, with the normal area highlighted in green.
when I asked them if there are more data that they want to see on the overview page, they showed interest on pressure, airway blockage and temperature. however, they also expressed the concern of information overload.
Thank you for reading til the end. If you are interested in the full paper, you can read it here https://ntnuopen.ntnu.no/ntnu-xmlui/handle/11250/2776862.
