The need for an integrated approach to remote monitoring of physiological data and activity data

The demographics of an ageing population who want to live independently and the increasing prevalence of chronic disease have been well documented. There is an associated call for research into ways of monitoring and managing these conditions. However research into remote monitoring the environment of people at home (variously called telecare, independent living or activities of daily living) and monitoring the physiological variables of patients at home (tele-monitoring, sometimes known as telehealth in the UK) has tended to be separate. Furthermore remote monitoring of activity data and remote monitoring of physiological data are often managed by separate organisations using separate systems provided by separate manufacturers, i.e. there are independent ‘‘silos’’ of management and care. In addition, there appear to be differences in the approach to the research. Research into remote monitoring of activity data generally aims to develop better sensors, while research into remote monitoring of physiological data usually aims to measure the benefit in outcomes from using existing technology. We recently conducted a systematic review of the literature on these two forms of remote monitoring to identify work where the two types of monitoring had been combined and there had been correlation of data sets. We were unable to identify any work that had been reported. In general the literature on remote monitoring of activity data only reported advances in the sensors and technology development. Testing of the sensors was restricted to normal subjects and only simple evaluations of usability had been undertaken with patients. There were no reports of physiological data observed at the same time, or recording of clinical events for correlation. There were no reports of the use of sensors for monitoring activities of daily living in the literature on remote monitoring of physiological data. This dearth of information exists despite anecdotal evidence that there are characteristic changes in the behaviour of patients with chronic disease when exacerbations occur. For example, patients exhibit a preference to sleep upright in a chair rather than prone in a bed with the onset of exacerbation in congestive heart failure; disturbed sleep patterns occur with a fall in blood oxygen in patients with chronic obstructive pulmonary disease (COPD). In our own work we have developed a system that integrates sensors to monitor activities of daily living (motion and bed/chair sensors) together with physiological sensors (blood pressure, blood glucose, SpO2, bodyweight) and we have observed changes in behaviour that correlate with changes in physical condition. As an example, Figure 1 shows a summary of night-time activity recorded by a Passive Infrared (PIR) motion sensor over a period of 21 weeks for a COPD patient. In the first half of the monitoring period the patient has a good sleep pattern, with little observed activity. However about halfway through the monitoring period this changes, and sleep appears to become disturbed, with much greater night time activity. The daily measurements of SpO2 for the same period show a marked fall in blood oxygen, which coincides with the change in sleep pattern. To date we have only been able to capture a few such related episodes and we do not yet have sufficient data to provide statistics on the accuracy of this technique to predict clinical events. In view of the well-known changes in behaviour that accompany deterioration in health in patients, it seems likely that simple unobtrusive monitoring of patients will be sufficient to indicate exacerbation. More research into gathering the necessary data to determine the accuracy of this integrated approach is required.