Effect of Home Telehealth Data Quality on Decision Support System Performance

Mas S. Mohktar, Jumadi Abd Sukor, Stephen J. Redmond, Jim Basilakis, Nigel H. Lovell

Abstract: In this study, data quality analyses were performed on raw signals from two types of home telehealth measurements; the pulse oximetry and the blood pressure. The results have confirmed that home telehealth pulse oximetry and blood pressure data quality issues do affect the reliability of a decision support system (DSS) for the particular algorithms and data sets used in this study. Both techniques (the manual outlier removal and the automated signal quality analysis) have improved the performance of the DSS. Therefore, these automated signal quality tools are considered useful and will be included in the DSS for the purpose of data quality assurance. This finding has also provided an additional method that can reduce the workload imposed when performing signal recording verification manually.