Society is gradually becoming more tuned-in to the way that equipment can be programmed to deliver a more personalised service – whether that’s retailers telling you what you might enjoy reading, listening to or prefer wearing; to your fitness wearable tracking your performance on your daily run. Our Digital Health Research programme is exploring how these approaches can be applied to CF care.
The aim is that in the future, this may mean fewer face-to-face clinic appointments, and faster diagnosis and more effective treatments when health problems develop. The ultimate goal is to develop a tailored, personalised health plan for each individual with cystic fibrosis.
Our digital health research programme began with an adult remote health monitoring study known as SmartCareCF in 2016. This was later followed by a programme exploring how remote health monitoring could be performed in children in our CLIMB-CF study. As well as providing adults and children with CF and childrens’ parents with ‘realtime’ data on how their CF is on a day-to-day level, the data from these studies are in the process of being analysed by computer scientists as part of our UK Cystic Fibrosis Innovation Hub at the University of Cambridge. The aim of their analysis is to be able to develop ways of predicting down turns in health known as lung exacerbations, and applying this knowledge to develop new models of care in our Project Breathe research study. The development of longer-term predictions of prognosis and tailored, personalised treatment plans is underway using data from the UK CF Registry. More detail on each of these studies is given below.
Testing the feasibility of adult health monitoring
Our interest in the use of remote monitoring for people with CF arose when Professor Andres Floto at the University of Cambridge first reported the outcome of a very small study in 2012. The Trust then explored with him expanding this ’feasibility’ study by supporting SmartCare CF.
148 adults with CF participated in the SmartCareCF study in seven CF centres across England led by Professor Andres Floto, where each person was asked to participate in the study for six months. The aim was to test the feasibility of people with CF monitoring their health at home, as well as generating data and samples for further analysis. The study involved using internet-enabled medical equipment such as spirometers, oxygen monitors and adapted weighing scales on a daily basis, where the information from the equipment was automatically pulled into a mobile handset. People were also asked to manually input a daily score for how well they felt and how their cough was that day. In addition to collecting these data, participants were asked to collect sputum samples on a daily basis for future biological analysis. Each participant had access to their own health data, to be able to track their own health over time, this data was not shared with their CF team and it didn’t affect the CF care they received. The data were also stored securely and anonymously for future research analysis.
Those who took part in SmartCareCF reported the value of seeing their own data and felt they had more control over their cystic fibrosis. Of the people who stopped the study early, some found that remote monitoring generated anxiety, which is completely understandable. At the Trust, we recognise that this approach will not suit everyone but believe in giving people options for how they receive their CF care. Sammie, a participant in the programme, explained her experience of the SmartCareCF in a video recorded for our UK CF Conference in 2019.
CLIMB-CF – assessing digital health in children
Alongside SmartCareCF we also funded a children’s home monitoring and study called CLIMB-CF, led by Professor Jane Davies and Dr Claire Edmondson at Imperial College London and Royal Brompton Hospital. A number of different factors needed to be considered before introducing digital health for children with CF in comparison to adults, so it was important to conduct a separate feasibility study in children. 144 children with CF and their parents took part in a similar six-month study, where the ages ranged from two years of age through to teenagers. The study was run in six areas of the UK, and CF Canada funded an extension of the study to two centres in Canada.
In comparison to the SmartCareCF adult study, a greater emphasis was placed on understanding any effects of regular home monitoring on the mental health and wellbeing for both the children with CF and their parents. Other differences included changes to the app for entering the data, which was co-designed with input from the study participants and the Trust’s Youth Advisory Group, and collecting some aspects of home monitoring data less frequently (several times a week, as opposed to daily collection). Analysis of this study is still ongoing, you can hear an update on the results so far from Professor Jane Davies and Dr Claire Edmondson at our UK CF Conference in 2019.
A continuous improvement cycle – refining the monitoring to improve health
The SmartCareCF study showed that asking adults with CF to monitor their health on a daily basis was feasible and it was found to be generally empowering. With a grant from the CF Foundation in USA, the next step towards digital health for people with CF is to test whether the home monitoring data can be used in real time to improve how people’s health is managed. In an ongoing study known as Project Breathe participants are sharing their data with clinicians, and using it to review whether a face-to-face clinic appointment is necessary, on an appointment-by-appointment basis. The study is aiming to recruit 400 adults with CF from three centres, initially in Cambridge (200 participants), and later in Cardiff (100 participants) and Glasgow (100 participants), where each person will participate in the study for 12 months.
As there have been advances in home monitoring technology since the original SmartCareCF study in 2016, and increasingly sophisticated phone app development; the technology has been updated to improve the experience of doing the home monitoring tests and a smartphone app for people to review their data has been designed.
During the COVID-19 pandemic the app was adapted so that everyone with CF can use it, rather than only the participants in the research study. The set-up of centres in Cardiff and Glasgow was also accelerated in light of the pandemic.
A continuous improvement cycle – analysing the data to improve the monitoring
The vast range of home monitoring data collected in the SmartCareCF study is currently being analysed within the UK Cystic Fibrosis Innovation Hub based at University of Cambridge to understand more about lung exacerbations that people with CF develop. Around once or twice a year, people with CF will experience a sudden worsening of their lung health known as an exacerbation, requiring intravenous antibiotics, long stays in hospital and causing disruption to day-to-day life. People with CF frequently report that they know an exacerbation is on the way, long before the symptoms lead to a drop in FEV1 and a hospital admission. SmartCareCF was designed to collect data at home to see if it was possible to pick up indications earlier using artificial intelligence.
If the onset of an exacerbation could be detected earlier, then it may be possible to improve the way it is managed with better long-term outcomes and less disruption to day-to-day life. Currently this early diagnosis is not possible due to a poor understanding of the early signs of an exacerbation and the underlying biological events that lead to it. The aim of one of the workstreams within the Innovation Hub is to address these gaps in our knowledge by analysing the home monitoring data and the sputum samples collected as part of the SmartCareCF study. The data also include information about when a participant in the study had an exacerbation.
Computer scientists working within the Innovation hub are analysing the changes in participants home monitoring data, before, during and after they had a lung exacerbation using machine learning algorithms. They are looking for patterns in the data that may predict when one is about to occur. So far, they’ve found that subtle changes in health monitoring measurements may predict when an exacerbation is likely to occur days before the symptoms develop. Dr John Winn who is working on this project gave an update on their analysis so far at the UK CF Conference. Work is underway to analyse the children’s home monitoring data from CLIMB-CF using similar methods.
Going forward anonymised data from participants in the Project Breathe study will feed into this data analysis, to help test these predictions. If the predictions are accurate, then it may change the way lung exacerbations are managed in the future.
Our principles of data sharing are at the heart of our digital health research programme. By making anonymised data widely accessible, we can maximise the impact of the research and improve the lives of all people with cystic fibrosis. We appreciate that many people have concerns about sharing their personal information, which is why this programme relies on the principled approach that people with CF own and are in control of their data. The data collected as part of these research studies is owned by the individual patients who collected it. People with CF can consent to share their data with their CF team. In time, we aim to be able to give people more choice regarding which data they share with their teams and any data they choose to withhold. We will strive to ensure that as many different monitoring devices as possible can be used by the digital health app we are developing. All final algorithms developed to predict health outcomes will be made available open access.
Future development plans include making all of the anonymised datasets available for researchers to use – on application and following robust governance and review procedures. We also plan to use the information within the datasets to enhance the UK CF Registry, ensuring a single, trusted database integrating data derived from people with cystic fibrosis.
Understanding my cystic fibrosis
To date, the data collected during home monitoring studies is a detailed look at a relatively short time period of the life of someone with CF to improve and manage their care in realtime. Professor Mihaela van der Schaar at the University of Cambridge is investigating whether analysis of data collected over a longer time period could identify when key changes in the health of someone with CF are likely to take place. For example, can she predict whether someone was likely to develop a specific complication of CF and if so at what age?
Artificial intelligence analysis of anonymised data from the UK CF Registry is being used to develop the predictive algorithm equations. In the future it is hoped that these the equations can be used to predict particular events during disease progression to inform and improve the clinical management of cystic fibrosis.
My personal CF prescription
As well as seeing lots of variations in which symptoms develop and how they progress, even people with the same genotype will have a different response to CFTR modulators. The genetic make-up of someone with CF determines some of this variation, along with socioeconomic and environmental factors. The second aim of Professor van der Schaar’s analysis is to use UK CF Registry data to better understand each person’s response to CFTR modulators. The result of these analyses can be combined with other studies underway to truly determine a person’s personalised CF prescription.
Professor van der Schaar’s research into to personalised CF planning are funded through the Trust’s personalised healthcare grant from the CFF foundation and a number of grants from the CF Trust.