The feasibility of using the clinical health tracker app and web-interface to assess surgical risk in elective surgical patients.

None listed

Summary The World Health Organisations International Classification of Function describes disability as the relationship between body function, activity levels and the level of participation. A patient’s level of disability has been shown to have adverse associations with surgical outcome. We have remotely measured patients’ activity and participation, continuously and objectively, using a cross platform smartphone application (app). The app reports step and Global Positioning System (GPS) data to a cloud database. Patients are generally accepting of this technology and show high levels of satisfaction with the process. We have used the data to generate meaningful outcomes for patients such as time spent at home and step counts. We have monitored patients prior to and following cardiothoracic surgery. These data will have a role in perioperative assessment. Currently perioperative assessment is performed using questionnaires and clinical assessment. These achieve a C-statistic for mortality prediction of 0.8-0.9. Machine learning techniques will enable the smartphone data to generate more accurate risk models for patients undergoing surgery and provide prediction for long-term patient-centred outcomes. Our objective is to assess the feasibility for a larger study. We intend to conduct a multicentre study enrolling 500 smartphone owning patients undergoing surgery at one of three centres in South Australia. Patients will receive a letter with instructions on how to install our app pre-operatively and be followed up for 90 days. We will report the time taken to achieve ethics and governance approval at each site and the number of patients who successfully activate the app at least seven days prior to surgery. We will report on the completeness of the step and GPS data from the individuals enrolled. We will use the data to power a larger study to predict days free from hospital at 90-days, time spent at home, daily step count and time spent inactive at 90-days. Plain English Statement Increasingly patients own smartphones. The multitude of sensors held in these devices allow for remote monitoring of patients. We believe that by using these data in a machine learning model, we will improve the prediction of risk of surgery, and also provide data on outcomes that are more meaningful to patients. We intend to assess the feasibility of using a smartphone app to monitor patients prior to and following surgery.

No related papers found yet.