Every year, around 90% of adults living with sickle cell disease (SCD) face at least one severe pain episode, or vaso-occlusive crisis (VOC), each with unique triggers, symptoms, and varying intensity. While the experience of pain is deeply personal, the common thread is clear: these crises are unpredictable,debilitating, and take a toll on every aspect of life.
When Sanius first began its work in the space, we had a strong theory that a lack of sleep held a significant link to VOCs, and while we did see some notable correlations between factors such as deep sleep and quality of life measures, the impact of other key variables around crises pushed sleep back from the forefront somewhat. From a personal perspective, I was recently in San Francisco with jetlag so intense that I was only sleeping 3 hours a day here and there, unable to work out thanks to the pain from the resulting VOC.
This is something that has been echoed in the feedback from our SCD community workshops, where despite some differences in terms of pain locations,sensations, and triggers, we found a common thread of stress, lack of sleep,and fatigue in the days leading up to a VOC. For many, this wasn't just an inconvenience – it was a critical warning sign that something was wrong.
One thing these events do have in common across the community is the immense impact on quality of life and clinical outcomes, with VOCs linked to 78% of emergency department visits, 95% of inpatient admissions, and frequent re admissions in environments where the care those with SCD receive is all too often sub optimal. The first step to tackling this for patients and providers alike is to truly understand the lead-up to a crisis in a way that helps us to start to predict, intervene,and prevent – something with each VOC I become more and more convinced could have something to do with sleep.
Our Early Insights in the Sickle Cell Sleep Space
Much of our early work involved a deep dive into real-time data captured from SCD populations using clinically validated wearable devices, uncovering strong links between disrupted sleep and quality of life (QoL), and with poor sleep emerging as a clear warning sign for deteriorating health outcomes. Driven both by firsthand patient experiences and by previously published studies that had identified sleep-related metrics as important players in patient well being, linking sleep disturbances to acute and chronic pain, QoL, overall physical health,functional disability, and clinical depression, our aim was to detail more specific sleep metrics and better describe these within SCD.
Based on these findings, we identified an early association between the amount of deep sleep and patient-reported QoL, captured through the EQ-5D-5L. Those recording lower percentages of deep sleep in turn were found to report significantly higher average QoL scores, suggesting longer periods of deep sleep may be linked to lower self-perceived wellbeing. Whether these poorer QoL scores may have seen individuals enter longer periods of deep sleep, or whether it was indeed the impact of excess deep sleep itself on wellbeing, our initial insights opened up new avenues for understanding how sleep disturbances could contribute to VOCs,and how improving sleep may help manage these painful episodes and overall health outcomes.
Despite this,our ongoing work into predictive modelling to develop an algorithm for early VOC detection, based on the real-time patient-generated data captured through wearables and an ePRO mobile app, had seen a stronger weighting of other key metrics in comparison to core sleep variables. Notably, our first round of advanced modelling critically flagged, much as anecdotal feedback from our SCD community had, fatigue as a central component in the onset of a VOC.
With growing patient numbers, data points, and the utilisation of new modelling frameworks to enhance flexibility, customisation, and feature addition within the model, our more recently presented work at the American Society of Haematology (ASH) 2024 saw not only an increase in the number of VOCs predicted – from 84% to 92% – but also a greater weighting for automated, wearable-derived metrics. Indeed, this highlighted the emergence of sleep metrics such as deep sleep proportion within the top 10 variables of algorithm importance, bringing with it a shift away from a reliance on measures that required high manual recording.
Alongside VOC prediction, our recent work has also explored how these metrics change in the week preceding and following a patient-reported hospitalisation, including comparisons to the day of hospitalisation itself and in comparison to their general, day-to-day baseline. QoL, as denoted by the EQ-5D-5L and associated visual analogue Health State scores, saw significant changes over the pre- and post-hospitalisation periods in a pattern supporting its use as a remotely monitored tool to track potential trends that suggest an individual’s health might be severely deteriorating. In line with this, variables linked to sleep quality, such as self-reported ‘Fatigue’ scores and wearable-captured nightly wake-up counts and duration, showed an inverse trend of increasing disturbances and severity over the pre- to post-hospitalisation period.
Similarly, insights captured through a patient workshop highlighted several key features of the prodromal stage, particularly around sleep and fatigue. In parallel with a number of wearable-derived findings, 67% of patients identified fatigue and sleep disturbances as prominent triggers for VOCs. Indeed, wearable data further revealed statistically significant changes in sleep metrics, with sleep heart rates showing a notable increase during the prodromal phase.
Patients also showed notable increases in wake-up durations, deeper sleep, and prolonged periods before fully waking, while experiencing a shorter time to actually fall asleep during the prodromal phase compared to baseline. Notably, while analysis identified key trends such as 55% of patients experiencing an increase in the number of nightly sleep disturbances, variation existed between participants in regards to the mean change from baseline to the prodromal period. As such,these findings further emphasise the utility of metrics remotely trackable by wearable data – providing core insights into the prodromal phase of VOCs, but critically reinforcing the need for personalised monitoring to support early intervention strategies.
Transforming Insights into Impact | What Could the Future of Sleep Research Look Like?
We all have an idea of traditional sleep studies, perhaps conducted in clinical settings where you are asked to sleep in unfamiliar environments, connected to multiple sensors and wires. While such studies can uncover valuable insights into sleep disorders, they can in turn come with limitations. Most importantly, the setting itself holds the risk of causing anxiety that disrupts natural sleep patterns, with resulting data that may not truly reflect a patient’s baseline sleep experience.
More recent methods have seen the growing implementation of home-based approaches, enabling patients to sleep in familiar environments. These typically involve simpler,more focused monitoring, including the use of wearable devices or portable sensors that track sleep stages, heart rate, oxygen levels, and respiratory patterns. Critically, these at-home studies are less intrusive and can be more comfortable for patients, as such offering a more natural reflection of their sleep patterns, with the increasing use of wearable technologies and mobile health apps allowing for real-time, continuous tracking of sleep quality in real-world settings – opening up new opportunities for patient empowerment and personalised care.
The growing use of both consumer-grade and clinically validated devices to track sleep patterns presents us with a unique opportunity to deepen our understanding of how sleep affects our health and how lifestyle and treatments may alter sleep. This shift could open the door to new, more effective approaches to managing health and wellbeing.
As such, and by moving away from the confines of the clinical setting and monitors that are typically deployed for data capture across more limited periods, it is becoming increasingly possible to gain a clearer understanding of how sleep interacts with daily life, in the real world. Ultimately, these advancements will enable the growth of tailored innovations driven by such insights in a way that is critically disease-agnostic.
Want to explore how sleep patterns shape health outcomes? Get in touch with us today here or through info@saniushealth.com. Let’s work together to transform the future of patient care.
