Using Wearable Devices to Detect Early Signs of Sepsis in Cancer Patients
Sepsis can develop quickly, particularly in people whose immune systems are weakened by cancer or its treatment. Fever, chills, confusion, breathlessness, a racing heart and unusual weakness may signal a medical emergency, yet early symptoms can be subtle or mistaken for chemotherapy side effects. Wearable technology offers a way to watch for changing patterns before a patient becomes critically unwell.
Smartwatches, patches and connected sensors can continuously record temperature, heart rate, breathing rate, oxygen saturation, movement and sleep. The opportunity is promising, but a device cannot diagnose sepsis by itself. Its value lies in combining timely physiological data with symptoms, pathology, treatment history and clinical judgement, creating a safer pathway from research evidence to urgent care.
Why Early Detection Matters In Cancer Care
Cancer patients may face several overlapping risks. Chemotherapy, stem cell transplantation, immunotherapy, surgery and some targeted treatments can reduce immune defences or damage protective barriers. A bloodstream infection may therefore progress rapidly, while the expected inflammatory response, such as a high fever, may be absent or delayed.
Sepsis screening traditionally relies on observations taken at intervals in hospital. Wearable sensors could add a continuous view, identifying a sustained rise in pulse, a gradual fall in oxygen levels or changes in respiratory rate. A single abnormal reading is rarely meaningful; a pattern across several hours may be more useful.
Any alert must be interpreted alongside the patient’s baseline. A person with lung cancer may have a lower usual oxygen saturation, while someone receiving steroids may show limited fever. Algorithms need to account for treatment cycles, comorbidities, pain, dehydration and anxiety rather than treating every deviation as infection.
Signals Worth Tracking
Wearable monitoring should focus on clinically relevant trends, with alerts designed to support nurses, doctors and patients rather than create alarm fatigue. Devices must be comfortable during sleep, simple to charge and reliable across different skin tones, body types and levels of digital confidence.
Research teams can begin by testing whether measurements are accurate enough for the intended setting, then assess whether an alert leads to faster review, antibiotics when appropriate or transfer to a higher level of care. The following signals may be useful when combined with symptoms and clinical assessment.
Physiological Changes
- Sustained increases in heart rate or breathing rate
- Falling oxygen saturation or skin temperature changes
- Reduced movement, sleep disruption or sudden fatigue
- New patterns of low blood pressure where suitable sensors are available
Patient-Reported Warning Signs
- Shivering, new confusion or feeling unusually faint
- Shortness of breath, worsening pain or clammy skin
- Reduced urine output, vomiting or difficulty drinking
- A strong sense that something is seriously wrong
From Hospital Monitoring To Home Support
The strongest use case may be the transition between hospital and home. Many people in Brisbane receive cancer treatment through large metropolitan services, then return to suburbs or regional communities. A connected device could send agreed observations to a clinical team during a high-risk period, helping staff decide whether to call the patient, arrange same-day assessment or direct them to emergency care.
This model must be practical for Australian households. Patients may have patchy mobile coverage while travelling through regional Queensland, limited internet access or difficulty using a smartphone after treatment. Services should provide clear instructions, loan equipment where needed and a telephone alternative. A wearable alert should never replace calling 000 when someone is severely breathless, confused, collapsed or rapidly deteriorating.
The patient’s care plan should state who reviews the data, during which hours and what happens when a threshold is crossed. Without this operating model, continuous monitoring can produce notifications that nobody is responsible for managing. In a busy Queensland hospital, a safe escalation pathway matters as much as the sensor itself.
Evidence, Privacy And Clinical Safety
Before widespread adoption, studies need to show more than improved data collection. They should measure time to clinical review, unplanned intensive care admission, antibiotic appropriateness, patient experience and outcomes for Aboriginal and Torres Strait Islander peoples, older adults and people living outside major cities. Researchers should also report false alarms, missed cases and differences between device brands.
Privacy is central because wearable data can reveal health status, location, sleep and daily routines. Patients need plain-English information about what is collected, where it is stored, who can access it and how long it is retained. Consent should cover research use separately from routine care, with a clear option to stop sharing without losing access to treatment.
Governance must include clinicians, cancer patients, carers, data specialists, Aboriginal health representatives and consumer advocates. Australian requirements may involve hospital privacy policies, human research ethics review, cybersecurity controls and Therapeutic Goods Administration considerations, depending on how the software and device are used. A prediction tool that influences urgent treatment requires stronger oversight than a general wellness tracker.
Translating Research Into Queensland Practice
Queensland has the scale to test this technology across tertiary hospitals, community services and regional pathways. A study could link cancer services at the Royal Brisbane and Women’s Hospital or Princess Alexandra Hospital with community oncology providers, pathology services and retrieval networks. It should also examine what happens after hours, when patients may present to an emergency department unfamiliar with their oncology history.
The Brisbane health network brings together health services, universities and research institutes to help move evidence into clinical practice. That kind of collaboration can support a study that connects engineering expertise with infection specialists, oncology teams, consumers and implementation researchers. It can also help define outcomes that matter beyond a device’s technical accuracy.
Local realities should shape the design. A patient in Toowoomba, Bundaberg or far north Queensland may need a different response from someone living close to a major Brisbane hospital. Telehealth, Medicare-funded care, private oncology clinics and public hospital pathways may intersect, so responsibilities and costs must be mapped before recruitment begins. Communication should use familiar language, including a clear instruction to “ring 000” for an emergency rather than relying on an app notification.
Successful translation will depend on trust. Patients are more likely to wear a sensor when it is comfortable, the purpose is explained honestly and the care team responds consistently. Clinicians are more likely to adopt it when alerts are validated, integrated into existing records and supported by adequate staffing. The goal is a responsive safety net that strengthens human care, not a substitute for examination and experienced judgement.
Health services, researchers and consumer partners can now work together to test wearable sepsis detection in carefully designed Australian studies. By combining continuous monitoring with equitable access, robust governance and rapid clinical escalation, Queensland can help turn promising digital health research into earlier recognition and safer care for people living with cancer.
