Prognostic value of heart rate variability for risk of serious adverse events in continuously monitored hospital patients.

Technological advances allow continuous vital sign monitoring at the general ward, but traditional vital signs alone may not predict serious adverse events (SAE). This study investigated continuous heart rate variability (HRV) monitoring's predictive value for SAEs in acute medical and major surgical patients. Data was collected from four prospective observational studies and two randomized controlled trials using a single-lead ECG. The primary outcome was any SAE, secondary outcomes included all-cause mortality and specific non-fatal SAE groups, all within 30 days. Subgroup analyses of medical and surgical patients were performed. The primary analysis compared the last 24 h preceding an SAE with the last 24 h of measurements in patients without an SAE. The area under a receiver operating characteristics curve (AUROC) quantified predictive performance, interpretated as low prognostic ability (0.5-0.7), moderate prognostic ability (0.7-0.9), or high prognostic ability (> 0.9). Of 1402 assessed patients, 923 were analysed, with 297 (32%) experiencing at least one SAE. The best performing threshold had an AUROC of 0.67 (95% confidence interval (CI) 0.63-0.71) for predicting cardiovascular SAEs. In the surgical subgroup, the best performing threshold had an AUROC of 0.70 (95% CI 0.60-0.81) for neurologic SAE prediction. In the medical subgroup, thresholds for all-cause mortality, cardiovascular, infectious, and neurologic SAEs had moderate prognostic ability, and the best performing threshold had an AUROC of 0.85 (95% CI 0.76-0.95) for predicting neurologic SAEs. Predicting SAEs based on the accumulated time below thresholds for individual continuously measured HRV parameters demonstrated overall low prognostic ability in high-risk hospitalized patients. Certain HRV thresholds had moderate prognostic ability for prediction of specific SAEs in the medical subgroup.

Genome-wide association studies with experimental validation identify a protective role for B lymphocytes against chronic post-surgical pain.

BACKGROUND: Chronic post-surgical pain (CPSP) significantly impacts patients' recovery and quality of life. Although environmental risk factors are well-established, genetic risk remains less understood. METHODS: A meta-analysis of genome-wide association studies followed by partitioned heritability was performed on 1350 individuals across five surgery types: hysterectomy, mastectomy, abdominal, hernia, and knee. In subsequent animal studies, withdrawal thresholds to evoked mechanical stimulation were measured in Rag1 null mutant and wild-type mice after plantar incision and laparotomy. Cell sorting by flow cytometry tracked recruitment of immune cell types. RESULTS: We discovered 77 genome-wide significant single-nucleotide polymorphism (SNP) hits, distributed among 24 loci and 244 genes. Meta-analysis of all cohorts estimated a SNP-based narrow-sense heritability for CPSP at ∼39%, indicating a substantial genetic contribution. Partitioned heritability analysis across a wide variety of tissues revealed enrichment of heritability in immune system-related genes, particularly those associated with B and T cells. Rag1 null mutant mice lacking both T and B cells exhibited exacerbated and prolonged allodynia up to 42 days after surgery, which was rescued by B-cell transfer. Recruitment patterns of B cells but not T cells differed significantly during the first 7 days after injury in the footpad, lymph nodes, and dorsal root ganglia. CONCLUSIONS: These findings suggest a key protective role for the adaptive immune system in the development of chronic post-surgical pain.

Excessive Oxygen Administration in High-Risk Patients Admitted to Medical and Surgical Wards Monitored by Wireless Pulse Oximeter.

The monitoring of oxygen therapy when patients are admitted to medical and surgical wards could be important because exposure to excessive oxygen administration (EOA) may have fatal consequences. We aimed to investigate the association between EOA, monitored by wireless pulse oximeter, and nonfatal serious adverse events (SAEs) and mortality within 30 days. We included patients in the Capital Region of Copenhagen between 2017 and 2018. Patients were hospitalized due to acute exacerbation of chronic obstructive pulmonary disease (AECOPD) or after major elective abdominal cancer surgery, and all were treated with oxygen supply. Patients were divided into groups by their exposure to EOA: no exposure, exposure for 1-59 min or exposure over 60 min. The primary outcome was SAEs or mortality within 30 days. We retrieved data from 567 patients for a total of 43,833 h, of whom, 63% were not exposed to EOA, 26% had EOA for 1-59 min and 11% had EOA for ≥60 min. Nonfatal SAEs or mortality within 30 days developed in 24%, 12% and 22%, respectively, and the adjusted odds ratio for this was 0.98 (95% CI, 0.96-1.01) for every 10 min. increase in EOA, without any subgroup effects. In conclusion, we did not observe higher frequencies of nonfatal SAEs or mortality within 30 days in patients exposed to excessive oxygen administration.

The future of postoperative vital sign monitoring in general wards: improving patient safety through continuous artificial intelligence-enabled alert formation and reduction.

PURPOSE: Monitoring of vital signs at the general ward with continuous assessments aided by artificial intelligence (AI) is increasingly being explored in the clinical setting. This review aims to describe current evidence for continuous vital sign monitoring (CVSM) with AI-based alerts - from sensor technology, through alert reduction, impact on complications, and to user-experience during implementation. RECENT FINDINGS: CVSM identifies significantly more vital sign deviations than manual intermittent monitoring. This results in high alert generation without AI-evaluation, both in patients with and without complications. Current AI is at the rule-based level, and this potentially reduces irrelevant alerts and identifies patients at need. AI-aided CVSM identifies complications earlier with reduced staff workload and a potential reduction of severe complications. SUMMARY: The current evidence for AI-aided CSVM suggest a significant role for the technology in reducing the constant 10-30% in-hospital risk of severe postoperative complications. However, large, randomized trials documenting the benefit for patient improvements are still sparse. And the clinical uptake of explainable AI to improve implementation needs investigation.

Deviating vital signs in continuous monitoring prior to discharge and risk of readmission: an observational study.

Premature discharge may result in readmission while longer hospitalization may increase risk of complications such as immobilization and reduce hospital capacity. Continuous monitoring detects more deviating vital signs than intermittent measurements and may help identify patients at risk of deterioration after discharge. We aimed to investigate the association between deviating vital signs detected by continuous monitoring prior to discharge and risk of readmission within 30 days. Patients undergoing elective major abdominal surgery or admitted with acute exacerbation of chronic obstructive pulmonary disease were included in this study. Eligible patients had vital signs monitored continuously within the last 24 h prior to discharge. The association between sustained deviated vital signs and readmission risk was analyzed by using Mann-Whitney's U test and Chi-square test. A total of 51 out of 265 patients (19%) were readmitted within 30 days. Deviated respiratory vital signs occurred frequently in both groups: desaturation < 88% for at least ten minutes was seen in 66% of patients who were readmitted and in 62% of those who were not (p = 0.62) while desaturation < 85% for at least five minutes was seen in 58% of readmitted and 52% of non-readmitted patients (p = 0.5). At least one sustained deviated vital sign was detected in 90% and 85% of readmitted patients and non-readmitted patients, respectively (p = 0.2). Deviating vital signs prior to hospital discharge were frequent but not associated with increased risk of readmission within 30 days. Further exploration of deviating vital signs using continuous monitoring is needed.

Quantifying physiological stability in the general ward using continuous vital signs monitoring: the circadian kernel density estimator.

Technological advances seen in recent years have introduced the possibility of changing the way hospitalized patients are monitored by abolishing the traditional track-and-trigger systems and implementing continuous monitoring using wearable biosensors. However, this new monitoring paradigm raise demand for novel ways of analyzing the data streams in real time. The aim of this study was to design a stability index using kernel density estimation (KDE) fitted to observations of physiological stability incorporating the patients' circadian rhythm. Continuous vital sign data was obtained from two observational studies with 491 postoperative patients and 200 patients with acute exacerbation of chronic obstructive pulmonary disease. We defined physiological stability as the last 24 h prior to discharge. We evaluated the model against periods of eight hours prior to events defined either as severe adverse events (SAE) or as a total score in the early warning score (EWS) protocol of ≥ 6, ≥ 8, or ≥ 10. The results found good discriminative properties between stable physiology and EWS-events (area under the receiver operating characteristics curve (AUROC): 0.772-0.993), but lower for the SAEs (AUROC: 0.594-0.611). The time of early warning for the EWS events were 2.8-5.5 h and 2.5 h for the SAEs. The results showed that for severe deviations in the vital signs, the circadian KDE model can alert multiple hours prior to deviations being noticed by the staff. Furthermore, the model shows good generalizability to another cohort and could be a simple way of continuously assessing patient deterioration in the general ward.

Wireless Single-Lead versus Standard 12-Lead ECG, for ST-Segment Deviation during Adenosine Cardiac Stress Scintigraphy.

Wearable wireless electrocardiographic (ECG) monitoring is well-proven for arrythmia detection, but ischemia detection accuracy is not well-described. We aimed to assess the agreement of ST-segment deviation from single- versus 12-lead ECG and their accuracy for the detection of reversible ischemia. Bias and limits of agreement (LoA) were calculated between maximum deviations in ST segments from single- and 12-lead ECG during 82Rb PET-myocardial cardiac stress scintigraphy. Sensitivity and specificity for reversible anterior-lateral myocardial ischemia detection were assessed for both ECG methods, using perfusion imaging results as a reference. Out of 110 patients included, 93 were analyzed. The maximum difference between single- and 12-lead ECG was seen in II (-0.019 mV). The widest LoA was seen in V5, with an upper LoA of 0.145 mV (0.118 to 0.172) and a lower LoA of -0.155 mV (-0.182 to -0.128). Ischemia was seen in 24 patients. Single-lead and 12-lead ECG both had poor accuracy for the detection of reversible anterolateral ischemia during the test: single-lead ECG had a sensitivity of 8.3% (1.0-27.0%) and specificity of 89.9% (80.2-95.8%), and 12-lead ECG a sensitivity of 12.5% (3.0-34.4%) and a specificity of 91.3% (82.0-96.7%). In conclusion, agreement was within predefined acceptable criteria for ST deviations, and both methods had high specificity but poor sensitivity for the detection of anterolateral reversible ischemia. Additional studies must confirm these results and their clinical relevance, especially in the light of the poor sensitivity for detecting reversible anterolateral cardiac ischemia.

Genome-wide association study suggests a critical contribution of the adaptive immune system to chronic post-surgical pain.

Chronic post-surgical pain affects a large proportion of people undergoing surgery, delaying recovery time and worsening quality of life. Although many environmental variables have been established as risk factors, less is known about genetic risk. To uncover genetic risk factors we performed genome-wide association studies in post-surgical cohorts of five surgery types- hysterectomy, mastectomy, abdominal, hernia, and knee- totaling 1350 individuals. Genetic associations between post-surgical chronic pain levels on a numeric rating scale (NRS) and additive genetic effects at common SNPs were evaluated. We observed genome-wide significant hits in almost all cohorts that displayed significance at the SNP, gene, and pathway levels. The cohorts were then combined via a GWAS meta-analysis framework for further analyses. Using partitioned heritability, we found that loci at genes specifically expressed in the immune system carried enriched heritability, especially genes related to B and T cells. The relevance of B cells in particular was then demonstrated in mouse postoperative pain assays. Taken altogether, our results suggest a role for the adaptive immune system in chronic post-surgical pain.

Associations between clinical interventions and transcutaneous blood gas values in postoperative patients.

PURPOSE: Postoperative monitoring of circulation and respiration is pivotal to guide intervention strategies and ensure patient outcomes. Transcutaneous blood gas monitoring (TCM) may allow for noninvasive assessment of changes in cardiopulmonary function after surgery, including a more direct assessment of local micro-perfusion and metabolism. To form the basis for studies assessing the clinical impact of TCM complication detection and goal-directed-therapy, we examined the association between clinical interventions in the postoperative period and changes in transcutaneous blood gasses. METHODS: Two-hundred adult patients who have had major surgery were enrolled prospectively and monitored with transcutaneous blood gas measurements (oxygen (TcPO2) and carbon dioxide (TcPCO2)) for 2 h in the post anaesthesia care unit, with recording of all clinical interventions. The primary outcome was changes in TcPO2, secondarily TcPCO2, from 5 min before a clinical intervention versus 5 min after, analysed with paired t-test. RESULTS: Data from 190 patients with 686 interventions were analysed. During clinical interventions, a mean change in TcPO2 of 0.99 mmHg (95% CI-1.79-0.2, p = 0.015) and TcPCO2 of-0.67 mmHg (95% CI 0.36-0.98, p < 0.001) was detected. CONCLUSION: Clinical interventions resulted in significant changes in transcutaneous oxygen and carbon dioxide. These findings suggest future studies to assess the clinical value of changes in transcutaneous PO2 and PCO2 in a postoperative setting. TRIAL REGISTRY: Clinical trial number: NCT04735380. CLINICAL TRIAL REGISTRY: https://clinicaltrials.gov/ct2/show/NCT04735380.

Continuous Glucose Monitoring Reveals Perioperative Hypoglycemia in Most Patients With Diabetes Undergoing Major Surgery: A Prospective Cohort Study.

OBJECTIVE: To investigate the frequency and duration of hypo- and hyperglycemia, assessed by continuous glucose monitoring (CGM) during and after major surgery, in departments with implemented diabetes care protocols. SUMMARY BACKGROUND DATA: Inadequate glycemic control in the perioperative period is associated with serious adverse events, but monitoring currently relies on point blood glucose measurements, which may underreport glucose excursions. METHODS: Adult patients without (A) or with diabetes [non-insulin-treated type 2 (B), insulin-treated type 2 (C) or type 1 (D)] undergoing major surgery were monitored using CGM (Dexcom G6), with an electrochemical sensor in the interstitial fluid, during surgery and for up to 10 days postoperatively. Patients and health care staff were blinded to CGM values, and glucose management adhered to the standard diabetes care protocol. Thirty-day postoperative serious adverse events were recorded. The primary outcome was duration of hypoglycemia (glucose <70 mg/dL). Clinicaltrials.gov: NCT04473001. RESULTS: Seventy patients were included, with a median observation time of 4.0 days. CGM was recorded in median 96% of the observation time. The median daily duration of hypoglycemia was 2.5 minutes without significant difference between the 4 groups (A-D). Hypoglycemic events lasting ≥15 minutes occurred in 43% of all patients and 70% of patients with type 1 diabetes. Patients with type 1 diabetes spent a median of 40% of the monitoring time in the normoglycemic range 70 to 180 mg/dL and 27% in the hyperglycemic range >250 mg/dL. Duration of preceding hypo- and hyperglycemia tended to be longer in patients with serious adverse events, compared with patients without events, but these were exploratory analyses. CONCLUSIONS: Significant duration of both hypo- and hyperglycemia was detected in high proportions of patients, particularly in patients with diabetes, despite protocolized perioperative diabetes management.

Haemodynamic and respiratory perioperative outcomes for open versus robot-assisted radical cystectomy: A double-blinded, randomised trial.

BACKGROUND: The clinical impact of prolonged steep Trendelenburg position and CO2 pneumoperitoneum during robot-assisted radical cystectomy (RC) on intraoperative conditions and immediate postoperative recovery remains to be assessed. The current study investigates intraoperative and immediate postoperative outcomes for open RC (ORC) versus robot-assisted RC with intracorporal urinary diversion (iRARC) in a blinded randomised trial. We hypothesised that ORC would result in a faster haemodynamic and respiratory post-anaesthesia care unit (PACU) recovery compared to iRARC. METHODS: This study is a predefined sub-analysis of a single-centre, double-blinded, randomised feasibility study. Fifty bladder cancer patients were randomly assigned to ORC (n = 25) or iRARC (n = 25). Patients, PACU staff, and ward personnel were blinded to the surgical technique. Both randomisation arms followed the same anaesthesiologic procedure, fluid treatment plan, and PACU care. The primary outcome was immediate postoperative recovery using a standardised PACU Discharge Criteria (PACU-DC) score. Secondary outcomes included respiration- and arterial O2 saturation scores as well as perioperative interventions and recordings. RESULTS: All patients underwent the allocated treatment. The total PACU-DC score was highest 6 h postoperatively with no difference in the total score between randomisation arms (p = 0.80). Both the ORC and iRARC groups maintained a mean respiration- and arterial O2 saturation score below 1 (out of 3) throughout PACU stay. The iRARC patients had significantly, but clinically acceptable, higher maximum airway pressure and arterial blood pressure, as well as lower minimum pH levels. The ORC group received significantly more opioids after extubation but marginally less analgesics in the PACU, compared to the iRARC group. CONCLUSIONS: A prolonged Trendelenburg position and CO2 pneumoperitoneum was well-tolerated during iRARC, and immediate postoperative recovery was similar for ORC and iRARC patients.

The impact of continuous wireless monitoring on adverse device effects in medical and surgical wards: a review of current evidence.

Novel technologies allow continuous wireless monitoring systems (CWMS) to measure vital signs and these systems might be favorable compared to intermittent monitoring regarding improving outcomes. However, device safety needs to be validated because uncertain evidence challenges the clinical implementation of CWMS. This review investigates the frequency of device-related adverse events in patients monitored with CWMS in general hospital wards. Systematic literature searches were conducted in PubMed and Embase. We included trials of adult patients in general hospital wards monitored with CWMS. Our primary outcome was the frequency of unanticipated serious adverse device effects (USADEs). Secondary outcomes were adverse device effects (ADEs) and serious adverse device effects (SADE). Data were extracted from eligible studies and descriptive statistics were applied to analyze the data. Seven studies were eligible for inclusion with a total of 1485 patients monitored by CWMS. Of these patients, 54 patients experienced ADEs (3.6%, 95% CI 2.8-4.7%) and no USADEs or SADEs were reported (0%, 95% CI 0-0.31%). The studies of the SensiumVitals® patch, the iThermonitor, and the ViSi Mobile® device reported 28 (9%), 25 (5%), and 1 (3%) ADEs, respectively. No ADEs were reported using the HealthPatch, WARD 24/7 system, or Coviden Alarm Management. Current evidence suggests that CWMS are safe to use but systematic reporting of all adverse device effects is warranted.

Agreement between standard and continuous wireless vital sign measurements after major abdominal surgery: a clinical comparison study.

Objective. Continuous wireless monitoring outside the post-anesthesia or intensive care units may enable early detection of patient deterioration, but good accuracy of measurements is required. We aimed to assess the agreement between vital signs recorded by standard and novel wireless devices in postoperative patients.Approach. In 20 patients admitted to the post-anesthesia care unit, we compared heart rate (HR), respiratory rate (RR), peripheral oxygen saturation (SpO2), and systolic and diastolic blood pressure (SBP and DBP) as paired data. The primary outcome measure was the agreement between standard wired and wireless monitoring, assessed by mean bias and 95% limits of agreement (LoA). LoA was considered acceptable for HR and PR, if within ±5 beats min-1(bpm), while RR, SpO2, and BP were deemed acceptable if within ±3 breaths min-1(brpm), ±3%-points, and ±10 mmHg, respectively.Main results.The mean bias between standard versus wireless monitoring was -0.85 bpm (LoA -6.2 to 4.5 bpm) for HR, -1.3 mmHg (LoA -19 to 17 mmHg) for standard versus wireless SBP, 2.9 mmHg (LoA -17 to 22) for standard versus wireless DBP, and 1.7% (LoA -1.4 mmHg to 4.8 mmHg) for SpO2, comparing standard versus wireless monitoring. The mean bias of arterial blood gas analysis versus wireless SpO2measurements was 0.02% (LoA -0.02% to 0.06%), while the mean bias of direct observation of RR compared to wireless measurements was 0.0 brpm (LoA -2.6 brpm to 2.6 brpm). 80% of all values compared were within predefined clinical limits.Significance.The agreement between wired and wireless HR, RR, and PR recordings in postoperative patients was acceptable, whereas the agreement for SpO2recordings (standard versus wireless) was borderline. Standard wired and wireless BP measurements may be used interchangeably in the clinical setting.

Prediction of Serious Adverse Events from Nighttime Vital Signs Values.

The period directly following surgery is critical for patients as they are at risk of infections and other types of complications, often summarized as severe adverse events (SAE). We hypothesize that impending complications might alter the circadian rhythm and, therefore, be detectable during the night before. We propose a SMOTE-enhanced XGBoost prediction model that classifies nighttime vital signs depending on whether they precede a serious adverse event or come from a patient that does not have a complication at all, based on data from 450 postoperative patients. The approach showed respectable results, producing a ROC-AUC score of 0.65 and an accuracy of 0.75. These findings demonstrate the need for further investigation.

[The future patient monitoring in the bed ward].

Current monitoring of vital signs in hospital wards rely on infrequent manual measurements. This narrative review describes how new wearable devices with artificial intelligence interpretation may overcome this challenge by providing nurses with continuous data without inducing alarm fatigue. Severe complications in non-ICU hospital wards are commonly preceded by vital sign deviations, and an ideal monitoring system must keep the patient in focus through user-friendly design and evidence-based alerts with high predictive abilities.

Impact of high-dose glucocorticoid on endothelial damage after liver resection - a double-blinded randomized substudy.

OBJECTIVES: Postoperative endothelial damage potentially results in increased vascular leakage, tissue edema and subsequent complications. The preventive effect of glucocorticoids on endothelial damage after surgery is sparsely described, including the relation between endothelial damage and the postoperative inflammatory response. Thus, we aimed to assess the preventive effect of high-dose glucocorticoids on postoperative endothelial damage, and the association between endothelial damage and inflammation after surgery. METHODS: This was a predefined substudy of a randomized double-blinded clinical trial of methylprednisolone 10 mg/kg (high dose) vs. dexamethasone 8 mg (low dose) in patients undergoing liver resection at Rigshospitalet, Copenhagen. In total 25 patients undergoing major liver resection (11 in the high-dose group and 14 in the low-dose group) were included. The primary outcome was changed in five endothelial biomarkers and the secondary outcome was changes in inflammation [C-reactive protein (CRP)] for the first three postoperative days. RESULTS: No statistically significant difference was found for any endothelial biomarkers postoperatively between the two groups (P > 0.15, for all). High-dose glucocorticoids significantly reduced CRP on day 3 compared to low-dose glucocorticoids [median difference on a postoperative day 3, 59.6 g/L, (84.2; 27.1), P < 0.002]. No significant correlation between endothelial damage and CRP levels was seen. CONCLUSIONS: No significant effect of high- vs. low-dose glucocorticoids on development in endothelial biomarkers after major liver resection was observed. High-dose glucocorticoids reduce the inflammatory response though without correlation to endothelial damage. Future studies should assess the clinical impact of increased endothelial biomarkers for clinical perioperative outcomes.

Impact of high-dose glucocorticoid on endothelial damage after liver resection - a double-blinded randomized substudy.

OBJECTIVES: Postoperative endothelial damage potentially results in increased vascular leakage, tissue edema and subsequent complications. The preventive effect of glucocorticoids on endothelial damage after surgery is sparsely described, including the relation between endothelial damage and the postoperative inflammatory response. Thus, we aimed to assess the preventive effect of high-dose glucocorticoids on postoperative endothelial damage, and the association between endothelial damage and inflammation after surgery. METHODS: This was a predefined substudy of a randomized double-blinded clinical trial of methylprednisolone 10 mg/kg (high dose) vs. dexamethasone 8 mg (low dose) in patients undergoing liver resection at Rigshospitalet, Copenhagen. In total 25 patients undergoing major liver resection (11 in the high-dose group and 14 in the low-dose group) were included. The primary outcome was changed in five endothelial biomarkers and the secondary outcome was changes in inflammation [C-reactive protein (CRP)] for the first three postoperative days. RESULTS: No statistically significant difference was found for any endothelial biomarkers postoperatively between the two groups (P > 0.15, for all). High-dose glucocorticoids significantly reduced CRP on day 3 compared to low-dose glucocorticoids [median difference on a postoperative day 3, 59.6 g/L, (84.2; 27.1), P < 0.002]. No significant correlation between endothelial damage and CRP levels was seen. CONCLUSIONS: No significant effect of high- vs. low-dose glucocorticoids on development in endothelial biomarkers after major liver resection was observed. High-dose glucocorticoids reduce the inflammatory response though without correlation to endothelial damage. Future studies should assess the clinical impact of increased endothelial biomarkers for clinical perioperative outcomes.

Clinical impact of vital sign abnormalities in patients admitted with acute exacerbation of chronic obstructive pulmonary disease: an observational study using continuous wireless monitoring.

Early detection of abnormal vital signs is critical for timely management of acute hospitalised patients and continuous monitoring may improve this. We aimed to assess the association between preceding vital sign abnormalities and serious adverse events (SAE) in patients hospitalised with acute exacerbation of chronic obstructive pulmonary disease (AECOPD). Two hundred patients' vital signs were wirelessly and continuously monitored with peripheral oxygen saturation, heart rate, and respiratory rate during the first 4 days after admission for AECOPD. Non-invasive blood pressure was also measured every 30-60 min. The primary outcome was occurrence of SAE according to international definitions within 30 days and physiological data were analysed for preceding vital sign abnormalities. Data were presented as the mean cumulative duration of vital sign abnormalities per 24 h and analysed using Wilcoxon rank sum test. SAE during ongoing continuous monitoring occurred in 50 patients (25%). Patients suffering SAE during the monitoring period had on average 455 min (SD 413) per 24 h of any preceding vital sign abnormality versus 292 min (SD 246) in patients without SAE, p = 0.08, mean difference 163 min [95% CI 61-265]. Mean duration of bradypnea (respiratory rate < 11 min-1) was 48 min (SD 173) compared with 30 min (SD 84) in patients without SAE, p = 0.01. In conclusion, the duration of physiological abnormalities was substantial in patients with AECOPD. There were no statistically significant differences between patients with and without SAE in the overall duration of preceding physiological abnormalities.Study registration: http://ClinicalTrials.gov (NCT03660501). Date of registration: Sept 6 2018.

Prediction of serious outcomes based on continuous vital sign monitoring of high-risk patients.

Continuous monitoring of high-risk patients and early prediction of severe outcomes is crucial to prevent avoidable deaths. Current clinical monitoring is primarily based on intermittent observation of vital signs and the early warning scores (EWS). The drawback is lack of time series dynamics and correlations among vital signs. This study presents an approach to real-time outcome prediction based on machine learning from continuous recording of vital signs. Systolic blood pressure, diastolic blood pressure, heart rate, pulse rate, respiration rate and peripheral blood oxygen saturation were continuously acquired by wearable devices from 292 post-operative high-risk patients. The outcomes from serious complications were evaluated based on review of patients' medical record. The descriptive statistics of vital signs and patient demographic information were used as features. Four machine learning models K-Nearest-Neighbors (KNN), Decision Trees (DT), Random Forest (RF), and Boosted Ensemble (BE) were trained and tested. In static evaluation, all four models had comparable prediction performance to that of the state of the art. In dynamic evaluation, the models trained from the static evaluation were tested with continuous data. RF and BE obtained the lower false positive rate (FPR) of 0.073 and 0.055 on no-outcome patients respectively. The four models KNN, DT, RF and BE had area under receiver operating characteristic curve (AUROC) of 0.62, 0.64, 0.65 and 0.64 respectively on outcome patients. RF was found to be optimal model with lower FPR on no-outcome patients and a higher AUROC on outcome patients. These findings are encouraging and indicate that additional investigations must focus on validating performance in a clinical setting before deployment of the real-time outcome prediction.