Martine J M Breteler1, Eline J KleinJan, Daan A J Dohmen, Luke P H Leenen, Richard van Hillegersberg, Jelle P Ruurda, Kim van Loon, Taco J Blokhuis, Cor J Kalkman. 1. From the Department of Anesthesiology, University Medical Center Utrecht, Utrecht University, Utrecht, The Netherlands (M.J.M.B., E.J.K., K.v.L., C.J.K.) FocusCura, Driebergen-Rijsenburg, The Netherlands (M.J.M.B., D.A.J.D.) Department of Technical Medicine, University of Twente, Enschede, The Netherlands (E.J.K.) Department of Trauma Surgery (L.P.H.L.) Department of Gastrointestinal and Oncologic Surgery (R.v.H., J.P.R.) University Medical Center Utrecht, Utrecht, The Netherlands (L.P.H.L.) Department of Surgery, Maastricht University Medical Center, Maastricht, The Netherlands (T.J.B.).
Abstract
BACKGROUND: Vital signs are usually recorded once every 8 h in patients at the hospital ward. Early signs of deterioration may therefore be missed. Wireless sensors have been developed that may capture patient deterioration earlier. The objective of this study was to determine whether two wearable patch sensors (SensiumVitals [Sensium Healthcare Ltd., United Kingdom] and HealthPatch [VitalConnect, USA]), a bed-based system (EarlySense [EarlySense Ltd., Israel]), and a patient-worn monitor (Masimo Radius-7 [Masimo Corporation, USA]) can reliably measure heart rate (HR) and respiratory rate (RR) continuously in patients recovering from major surgery. METHODS: In an observational method comparison study, HR and RR of high-risk surgical patients admitted to a step-down unit were simultaneously recorded with the devices under test and compared with an intensive care unit-grade monitoring system (XPREZZON [Spacelabs Healthcare, USA]) until transition to the ward. Outcome measures were 95% limits of agreement and bias. Clarke Error Grid analysis was performed to assess the ability to assist with correct treatment decisions. In addition, data loss and duration of data gaps were analyzed. RESULTS: Twenty-five high-risk surgical patients were included. More than 700 h of data were available for analysis. For HR, bias and limits of agreement were 1.0 (-6.3, 8.4), 1.3 (-0.5, 3.3), -1.4 (-5.1, 2.3), and -0.4 (-4.0, 3.1) for SensiumVitals, HealthPatch, EarlySense, and Masimo, respectively. For RR, these values were -0.8 (-7.4, 5.6), 0.4 (-3.9, 4.7), and 0.2 (-4.7, 4.4) respectively. HealthPatch overestimated RR, with a bias of 4.4 (limits: -4.4 to 13.3) breaths/minute. Data loss from wireless transmission varied from 13% (83 of 633 h) to 34% (122 of 360 h) for RR and 6% (47 of 727 h) to 27% (182 of 664 h) for HR. CONCLUSIONS: All sensors were highly accurate for HR. For RR, the EarlySense, SensiumVitals sensor, and Masimo Radius-7 were reasonably accurate for RR. The accuracy for RR of the HealthPatch sensor was outside acceptable limits. Trend monitoring with wearable sensors could be valuable to timely detect patient deterioration.
BACKGROUND: Vital signs are usually recorded once every 8 h in patients at the hospital ward. Early signs of deterioration may therefore be missed. Wireless sensors have been developed that may capture patient deterioration earlier. The objective of this study was to determine whether two wearable patch sensors (SensiumVitals [Sensium Healthcare Ltd., United Kingdom] and HealthPatch [VitalConnect, USA]), a bed-based system (EarlySense [EarlySense Ltd., Israel]), and a patient-worn monitor (Masimo Radius-7 [Masimo Corporation, USA]) can reliably measure heart rate (HR) and respiratory rate (RR) continuously in patients recovering from major surgery. METHODS: In an observational method comparison study, HR and RR of high-risk surgical patients admitted to a step-down unit were simultaneously recorded with the devices under test and compared with an intensive care unit-grade monitoring system (XPREZZON [Spacelabs Healthcare, USA]) until transition to the ward. Outcome measures were 95% limits of agreement and bias. Clarke Error Grid analysis was performed to assess the ability to assist with correct treatment decisions. In addition, data loss and duration of data gaps were analyzed. RESULTS: Twenty-five high-risk surgical patients were included. More than 700 h of data were available for analysis. For HR, bias and limits of agreement were 1.0 (-6.3, 8.4), 1.3 (-0.5, 3.3), -1.4 (-5.1, 2.3), and -0.4 (-4.0, 3.1) for SensiumVitals, HealthPatch, EarlySense, and Masimo, respectively. For RR, these values were -0.8 (-7.4, 5.6), 0.4 (-3.9, 4.7), and 0.2 (-4.7, 4.4) respectively. HealthPatch overestimated RR, with a bias of 4.4 (limits: -4.4 to 13.3) breaths/minute. Data loss from wireless transmission varied from 13% (83 of 633 h) to 34% (122 of 360 h) for RR and 6% (47 of 727 h) to 27% (182 of 664 h) for HR. CONCLUSIONS: All sensors were highly accurate for HR. For RR, the EarlySense, SensiumVitals sensor, and Masimo Radius-7 were reasonably accurate for RR. The accuracy for RR of the HealthPatch sensor was outside acceptable limits. Trend monitoring with wearable sensors could be valuable to timely detect patient deterioration.
Authors: Jobbe P L Leenen; Eline M Dijkman; Joris D van Dijk; Henderik L van Westreenen; Cor Kalkman; Lisette Schoonhoven; Gijsbert A Patijn Journal: BMJ Open Date: 2021-02-17 Impact factor: 2.692
Authors: Morgan Le Guen; Pierre Squara; Sabrina Ma; Shérifa Adjavon; Bernard Trillat; Messaouda Merzoug; Philippe Aegerter; Marc Fischler Journal: BMJ Open Date: 2020-09-25 Impact factor: 2.692