The Measurement of VItal SIgns by Lifelight® Software in comparisON to the Standard of - Multi-site Development: The VISION-MD Study
试验速览
- 阶段
- 不适用
- 状态
- 已完成
- 发起方
- 入组人数
- 1,869
- 试验地点
- 1
- 主要终点
- Train Lifelight algorithms across more extensive clinical ranges, including patients with a systolic blood pressure of <90 or >180 mmHg, to progress towards British Hypertension Society Grade C and a standard deviation performance target of <8mmHg.
研究概览
简要总结
AIM: To advance the development and accuracy of the Lifelight® app for the measurement of vital signs, therefore developing a non-invasive and easy-to-perform means of measuring vital signs which can be implemented across a wide range of settings, both within hospitals and out in the community.
METHOD: Lifelight® is a computer program ("app") for measuring vital signs which can be used on smart devices that contain a camera. It is able to measure all of the vital signs by measuring very small changes in skin colour that occur each time the heart beats. This means that it does not need to touch the patient. The investigators believe this could be an effective way of measuring vital signs, especially during the COVID-19 pandemic when prevention of cross-contamination between patients is essential. Patients are also likely to be reassured by a contactless approach. The app uses data from looking at a person's face to calculate the vital signs. This is possible because there are tiny changes in facial skin that occur each time the heart beats.
The investigators believe Lifelight® could be an effective way of measuring vital signs. The app is still under development, which means that it is still "learning" the best match between the information it collects from the face and the values of vital signs measured using the standard equipment. The app should become more accurate in calculating the vital signs as it sees more and more information from patients.
So far, the app has seen data from inpatients, outpatients, patients attending GP surgeries and healthy people. This has improved its accuracy in measuring vital signs. However, the app needs to see more information so that it can be sufficiently accurate for specific clinical applications such as long-term monitoring of hypertension. To do this, it particularly needs to see information from people with abnormal blood pressures and blood oxygen levels. In order to capture the full range of observations, the app will need to be trialled with some of the most critically ill patients - some of these will not have capacity to consent to participation in the study. It also needs to see more data from people with different skin tones so the investigators can be sure it is accurate for all patients.
To do this, the investigators will recruit people who are attending one of two hospitals, either as an inpatient, an outpatient, a friend/relative of a patient, or a member of hospital staff. The exact number will depend on how quickly the app "learns" and how many of the vital signs are outside of the normal range. The investigators will take the participant's vital signs using standard clinical equipment whilst recording a video of their face. The investigators will use most of these measurements and video to teach the app how to become more accurate at measuring vital signs. The investigators will keep the remaining data separate and use it to test how accurate the app is. All of the data will be kept securely.
The investigators will also collect feedback from participants and healthcare staff on their experiences using the app and information that allows us to assess whether there are any savings to the healthcare economy through use of this technology.
详细描述
Following informed consent, the study staff member will complete a very brief set of demographic and medical history questions, limited to the presence or absence of medical problems and treatment for them.
Participants with capacity to consent will be recruited into either Sub-protocol 1, 2 or 3 on the basis of which vital signs are expected to be abnormal. Participants who lack capacity to consent to take part in the study will be recruited into Sub-protocol 4.
- Sub-protocol 1 participants will have blood pressure, oxygen saturation and heart rate measured.
- Sub-protocol 2 participants will have respiratory rate and oxygen saturation measured.
- Sub-protocol 3 participants will have oxygen saturation measured.
- Sub-protocol 4 participants will have blood pressure, heart rate, respiratory rate and oxygen saturation measured.
Participants may also be recruited to a sub-protocol on the basis of their skin tone. This is because there are targets within Sub-protocols 1, 2 and 3 related to skin tone.
Not all vital signs are collected in all participants to focus study nurse attention on fewer tasks and to avoid collection of data that is not subsequently used to meet the study objectives. This approach should help to keep all aspects of data collection as high quality as possible and is consistent with the GDPR requirement of data minimisation.
研究设计
- 研究类型
- Observational
- 观察模型
- Cohort
- 时间视角
- Prospective
入排标准
- 年龄范围
- 16 Years 至 —(Child, Adult, Older Adult)
- 性别
- All
- 接受健康志愿者
- 是
入选标准
- •16 years old or above
排除标准
- 未提供
结局指标
主要结局
Train Lifelight algorithms across more extensive clinical ranges, including patients with a systolic blood pressure of <90 or >180 mmHg, to progress towards British Hypertension Society Grade C and a standard deviation performance target of <8mmHg.
时间窗: 12 months
Train Lifelight algorithms on a more diverse population, including patients with Fitzpatrick scale 1, 4, 5 and 6 skin tones.
时间窗: 12 months
Train Lifelight algorithms across more extensive clinical ranges, including patients with a diastolic blood pressure of >110 mmHg, to improve Lifelight's British Hypertension Society Grade B grading and standard deviation performance target of <8mmHg
时间窗: 12 months
Train Lifelight algorithms across more extensive clinical ranges, including patients with a respiration rate of >25 rpm, to reach closer to the indicative performance target of a 5 breath per minute maximum error tolerance.
时间窗: 12 months
Train Lifelight algorithms across more extensive clinical ranges, including patients with a pulse of <40bpm or >120bpm, to reach closer to the indicative performance target of an RMSE of 3bpm or less.
时间窗: 12 months
Train Lifelight algorithms across more extensive clinical ranges, including patients with an blood oxygen saturation of <92%, to reach closer to a performance target of 4% maximum error tolerance.
时间窗: 12 months
次要结局
未报告次要终点
