A Comparison of 5 Algorithmic Methods and Machine Learning Pattern Recognition for Artifact Detection in Electronic Records of 5 Different Vital Signs: A Retrospective Analysis

BACKGROUND

Research on electronic health record physiological data is common, invariably including artifacts. Traditionally, these artifacts have been handled using simple filter techniques. The authors hypothesized different artifact detection algorithms, including machine learning, may be necessary to provide optimal performance for various vital signs and clinical contexts.

MATERIALS AND METHODS

In a retrospective single center study, intraoperative OR and ICU electronic health record datasets including heart rate, oxygen saturation, blood pressure, temperature, and capnometry were included. All records were screened for artifacts by at least two human experts. Classical artifact detection methods (cutoff, multiples of standard deviation (z-value), interquartile range, and local outlier factor) and a supervised learning model implementing long short-term memory neural networks were tested for each vital sign against the human expert reference dataset. For each artifact detection algorithm, sensitivity and specificity were calculated.

RESULTS

A total of 106 (53 operating room and 53 ICU) patients were randomly selected, resulting in 392,808 data points. Human experts annotated 5,167 (1.3%) data points as artifacts. The artifact detection algorithms demonstrated large variations in performance. The specificity was above 90% for all detection methods and all vital signs. The neural network showed significantly higher sensitivities than the classic methods for: heart rate (ICU: 33.6%, 95% CI: 33.1-44.6), systolic invasive blood pressure (both in the OR (62.2%, 95% CI: 57.5-71.9) and ICU (60.7%, 95% CI: 57.3-71.8), and temperature in the OR (76.1%, 95% CI: 63.6-89.7). The confidence intervals for specificity overlapped for all methods. Generally, sensitivity was low, with only the z-value for oxygen saturation in the operating room reaching 88.9%. All other sensitivities were less than 80%.

CONCLUSION

No single artifact detection method consistently performed well across different vital signs and clinical settings. Neural networks may be a promising artifact detection method for specific vital signs.

Profiling endogenous adrenal function during veno-venous ECMO support in COVID-19 ARDS: a descriptive analysis

Background

Prolonged critical illness is often accompanied by an impairment of adrenal function, which has been frequently related to conditions complicating patient management. The presumed connection between hypoxia and the pathogenesis of this critical- illness- related corticosteroid insufficiency (CIRCI) might play an important role in patients with severe acute respiratory distress syndrome (ARDS). Since extracorporeal membrane oxygenation (ECMO) is frequently used in ARDS, but data on CIRCI during this condition are scarce, this study reports the behaviour of adrenal function parameters during oxygenation support with veno-venous (vv)ECMO in coronavirus disease 2019 (COVID-19) ARDS.

Methods

A total of 11 patients undergoing vvECMO due to COVID-19 ARDS at the Medical University of Vienna, who received no concurrent corticosteroid therapy, were retrospectively included in this study. We analysed the concentrations of cortisol, aldosterone, and angiotensin (Ang) metabolites (Ang I-IV, Ang 1-7, and Ang 1-5) in serum via liquid chromatography/tandem mass spectrometry before, after 1 day, 1 week, and 2 weeks during vvECMO support and conducted correlation analyses between cortisol and parameters of disease severity.

Results

Cortisol concentrations appeared to be lowest after initiation of ECMO and progressively increased throughout the study period. Higher concentrations were related to disease severity and correlated markedly with interleukin-6, procalcitonin, pH, base excess, and albumin during the first day of ECMO. Fair correlations during the first day could be observed with calcium, duration of critical illness, and ECMO gas flow. Angiotensin metabolite concentrations were available in a subset of patients and indicated a more homogenous aldosterone response to plasma renin activity after 1 week of ECMO support.

Conclusion

Oxygenation support through vvECMO may lead to a partial recovery of adrenal function over time. In homogenous patient collectives, this novel approach might help to further determine the importance of adrenal stress response in ECMO and the influence of oxygenation support on CIRCI.

Accuracy of a Dual-Sensor Heat-Flux (DHF) Non-Invasive Core Temperature Sensor in Pediatric Patients Undergoing Surgery

Accurate temperature measurement is crucial for the perioperative management of pediatric patients, and non-invasive thermometry is necessary when invasive methods are infeasible. A prospective observational study was conducted on 57 patients undergoing elective surgery. Temperatures were measured using a dual-sensor heat-flux (DHF) thermometer (Tcore™) and a rectal temperature probe (TRec), and the agreement between the two measurements was assessed. The DHF measurements showed a bias of +0.413 °C compared with those of the TRec. The limits of agreement were broader than the pre-defined ±0.5 °C range (-0.741 °C and +1.567 °C). Although the DHF sensors tended to overestimate the core temperature compared to the rectal measurements, an error grid analysis demonstrated that 95.81% of the DHF measurements would not have led to a wrong clinical decision, e.g., warming or cooling when not necessary. In conclusion, the low number of measurements that would have led to incorrect decisions suggests that the DHF sensor can be considered an option for continuous temperature measurement when more invasive methods are infeasible.

Obesity and its influence on liver dysfunction, morbidity and mortality after liver resection

Background

Obesity and associated steatosis is an increasing health problem worldwide. Its influence on post-hepatectomy liver failure (PHLF) and after liver resection (LR) is still unclear.

Methods

Patients who underwent LR were investigated and divided into three groups [normal weight: body mass index (BMI) 18.5-24.9 kg/m2, overweight: BMI 25.0-29.9 kg/m2, obese: BMI ≥30 kg/m2] in this retrospective study. Primary aim of this study was to assess the influence of BMI and nonalcoholic steatohepatitis (NASH) on PHLF and morbidity.

Results

Of 888 included patients, 361 (40.7%) had normal weight, 360 (40.5%) were overweight, 167 (18.8%) were obese. Median age was 62.5 years (IQR, 54-69 years). The primary indication for LR was colorectal liver metastases (CLM) (n=366, 41.2%). NASH was present in 58 (16.1%) of normal weight, 84 (23.3%) of overweight and 69 (41.3%) of obese patients (P<0.001). PHLF occurred in 16.3% in normal weight, 15.3% in overweight and 11.4% in obese patients (P=0.32). NASH was not associated with PHLF. There was no association between patients' weight and the occurrence of postoperative complications (P=0.45). At multivariable analysis, solely major LR [odds ratio (OR): 2.7, 95% confidence interval (CI): 1.83-4.04; P<0.001] remained a significant predictor for PHLF.

Conclusions

Postoperative complications and PHLF are comparable in normal weight, overweight and obese patients and LRs using modern techniques can be safely performed in these patients.

The influence of explainable vs non-explainable clinical decision support systems on rapid triage decisions: a mixed methods study

BACKGROUND

During the COVID-19 pandemic, a variety of clinical decision support systems (CDSS) were developed to aid patient triage. However, research focusing on the interaction between decision support systems and human experts is lacking.

METHODS

Thirty-two physicians were recruited to rate the survival probability of 59 critically ill patients by means of chart review. Subsequently, one of two artificial intelligence systems advised the physician of a computed survival probability. However, only one of these systems explained the reasons behind its decision-making. In the third step, physicians reviewed the chart once again to determine the final survival probability rating. We hypothesized that an explaining system would exhibit a higher impact on the physicians' second rating (i.e., higher weight-on-advice).

RESULTS

The survival probability rating given by the physician after receiving advice from the clinical decision support system was a median of 4 percentage points closer to the advice than the initial rating. Weight-on-advice was not significantly different (p = 0.115) between the two systems (with vs without explanation for its decision). Additionally, weight-on-advice showed no difference according to time of day or between board-qualified and not yet board-qualified physicians. Self-reported post-experiment overall trust was awarded a median of 4 out of 10 points. When asked after the conclusion of the experiment, overall trust was 5.5/10 (non-explaining median 4 (IQR 3.5-5.5), explaining median 7 (IQR 5.5-7.5), p = 0.007).

CONCLUSIONS

Although overall trust in the models was low, the median (IQR) weight-on-advice was high (0.33 (0.0-0.56)) and in line with published literature on expert advice. In contrast to the hypothesis, weight-on-advice was comparable between the explaining and non-explaining systems. In 30% of cases, weight-on-advice was 0, meaning the physician did not change their rating. The median of the remaining weight-on-advice values was 50%, suggesting that physicians either dismissed the recommendation or employed a "meeting halfway" approach. Newer technologies, such as clinical reasoning systems, may be able to augment the decision process rather than simply presenting unexplained bias.

Inadequate Energy Delivery Is Frequent among COVID-19 Patients Requiring ECMO Support and Associated with Increased ICU Mortality

BACKGROUND

Patients receiving extracorporeal membrane oxygenation (ECMO) support are at high risk for malnutrition. There are currently no general nutrition guidelines for coronavirus disease 2019 (COVID-19) patients during ECMO therapy.

METHODS

We conducted a retrospective analysis of COVID-19 patients requiring venovenous ECMO support at a large tertiary hospital center. Nutrition goals were calculated using 25 kcal/kg body weight (BW)/day. Associations between nutrition support and outcome were evaluated using Kaplan-Meier and multivariable Cox regression analyses.

RESULTS

Overall, 102 patients accounted for a total of 2344 nutrition support days during ECMO therapy. On 40.6% of these days, nutrition goals were met. Undernutrition was found in 40.8%. Mean daily calorie delivery was 73.7% of calculated requirements, mean daily protein delivery was 0.7 g/kg BW/d. Mean energy intake of ≥70% of calculated targets was associated with significantly lower ICU mortality independently of age, disease severity at ECMO start and body mass index (adjusted hazard ratio: 0.372, p = 0.007).

CONCLUSIONS

Patients with a mean energy delivery of ≥70% of calculated targets during ECMO therapy had a better ICU survival compared to patients with unmet energy goals. These results indicate that adequate nutritional support needs to be a major priority in the treatment of COVID-19 patients requiring ECMO support.

Use of Recombinant Activated Factor VII in Bleeding Lung Transplant Patients Undergoing Perioperative ECMO Therapy

BACKGROUND

Hemostasis in critically ill patients represents a fragile balance between hypocoagulation and hypercoagulation, and is influenced by various factors. Perioperative use of extracorporeal membrane oxygenation (ECMO)-increasingly used in lung transplantation-further destabilizes this balance, not least due to systemic anticoagulation. In the case of massive hemorrhage, guidelines recommend considering recombinant activated Factor VII (rFVIIa) as an ultima ratio treatment only after several preconditions of hemostasis have been established. These conditions are calcium levels ≥ 0.9 mmol/L, fibrinogen levels ≥ 1.5 g/L, hematocrit ≥ 24%, platelet count ≥ 50 G/L, core body temperature ≥ 35 °C, and pH ≥ 7.2.

OBJECTIVES

This is the first study to examine the effect of rFVIIa on bleeding lung transplant patients undergoing ECMO therapy. The fulfillment of guideline-recommended preconditions prior to the administration of rFVIIa and its efficacy alongside the incidence of thromboembolic events were investigated.

METHODS

In a high-volume lung transplant center, all lung transplant recipients receiving rFVIIa during ECMO therapy between 2013 and 2020 were screened for the effect of rFVIIa on hemorrhage, fulfillment of recommended preconditions, and incidence of thromboembolic events.

RESULTS AND DISCUSSION

Of the 17 patients who received 50 doses of rFVIIa, bleeding ceased in four patients without surgical intervention. Only 14% of rFVIIa administrations resulted in hemorrhage control, whereas 71% of patients required revision surgery for bleeding control. Overall, 84% of all recommended preconditions were fulfilled; however, fulfillment was not associated with rFVIIa efficacy. The incidence of thromboembolic events within five days of rFVIIa administration was comparable to cohorts not receiving rFVIIa.

Progressive cholestasis and associated sclerosing cholangitis are frequent complications of COVID-19 in patients with chronic liver disease

BACKGROUND AND AIMS

Cholestasis is associated with disease severity and worse outcome in COVID-19. Cases of secondary sclerosing cholangitis (SSC) after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection have been described.

APPROACH AND RESULTS

Hospitalized patients with COVID-19 between 03/2020 and 07/2021 were included. Patients were stratified as having (i) no chronic liver disease (CLD), (ii) non-advanced CLD (non-ACLD), or (iii) advanced CLD (ACLD). Patients with CLD and non-COVID-19 pneumonia were matched to patients with CLD and COVID-19 as a control cohort. Liver chemistries before (Pre) and at first, second, and third blood withdrawal after SARS-CoV-2 infection (T1-T3) and at last available time point (last) were recorded. A total of 496 patients were included. In total, 13.1% (n = 65) had CLD (non-ACLD: 70.8%; ACLD: 29.2%); the predominant etiology was NAFLD/NASH (60.0%). COVID-19-related liver injury was more common among patients with CLD (24.6% vs. 10.6%; p = 0.001). After SARS-CoV-2 infection, patients with CLD exhibited progressive cholestasis with persistently increasing levels of alkaline phosphatase (Pre: 91.0 vs. T1: 121.0 vs. last: 175.0 U/L; p < 0.001) and gamma-glutamyl transferase (Pre: 95.0 vs. T1: 135.0 vs. last: 202.0 U/L; p = 0.001). A total of 23.1% of patients with CLD (n = 15/65) developed cholestatic liver failure (cholestasis plus bilirubin ≥6 mg/dl) during COVID-19, and 15.4% of patients (n = 10/65) developed SSC. SSC was significantly more frequent among patients with CLD and COVID-19 than in patients with CLD and non-COVID-19 pneumonia (p = 0.040). COVID-19-associated SSC occurred predominantly in patients with NAFLD/NASH and metabolic risk factors. A total of 26.3% (n = 5/19) of patients with ACLD experienced hepatic decompensation after SARS-CoV-2 infection.

CONCLUSIONS

About 20% of patients with CLD develop progressive cholestasis after SARS-CoV-2 infection. Patients with NAFLD/NASH and metabolic risk factors are at particular risk for developing cholestatic liver failure and/or SSC after COVID-19.

Subcutaneous Enoxaparin for Systemic Anticoagulation of COVID-19 Patients During Extracorporeal Life Support

Background

Extracorporeal membrane oxygenation, with an inherent requirement for anticoagulation to avoid circuit thrombosis, is a key element in the treatment of respiratory failure associated with COVID-19. Anticoagulation remains challenging, the standard of care being intravenous continuous administration of unfractionated heparin. Yet regimens vary. Some intensive care units in our center have successfully used enoxaparin subcutaneously in recent years and throughout the pandemic.

Methods

We retrospectively analyzed adult COVID-19 patients with respiratory failure who had been systemically anticoagulated using either enoxaparin or unfractionated heparin. The choice of anticoagulant therapy was based on the standard of the intensive care unit. Defined thromboembolic and hemorrhagic events were analyzed as study endpoints.

Results

Of 98 patients, 62 had received enoxaparin and 36 unfractionated heparin. All hazard ratios for the thromboembolic (3.43; 95% CI: 1.08–10.87; p = 0.04), hemorrhagic (2.58; 95% CI: 1.03–6.48; p = 0.04), and composite (2.86; 95% CI: 1.41–5.92; p = 0.007) endpoints favored enoxaparin, whose efficient administration was verified by peak levels of anti-factor Xa (median: 0.45 IU ml⁻¹; IQR: 0.38; 0.56). Activated partial thromboplastin time as well as thrombin time differed significantly (both p<0.001) between groups mirroring the effect of unfractionated heparin.

Conclusions

This study demonstrates the successful use of subcutaneous enoxaparin for systemic anticoagulation in patients with COVID-19 during extracorporeal membrane oxygenation. Our findings are to be confirmed by future prospective, randomized, controlled trials.

Age-adjusted mortality and predictive value of liver chemistries in a Viennese cohort of COVID-19 patients

BACKGROUND AND AIMS

The coronavirus disease of 2019 (COVID-19) causes considerable mortality worldwide. We aimed to investigate the frequency and predictive role of abnormal liver chemistries in different age groups.

METHODS

Patients with positive severe acute respiratory distress syndrome-coronavirus-2 (SARS-CoV-2) polymerase chain reaction (PCR) test between 03/2020-07/2021 at the Vienna General Hospital were included. Patients were stratified for age: 18-39 vs. 40-69 vs. ≥70 years (y). Aspartate aminotransferase (AST), alanine-aminotransferase (ALT), alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT) and total bilirubin (BIL) were recorded.

RESULTS

900 patients (18-39 years: 32.2%, 40-69 years: 39.7%, ≥70 years: 28.1%) were included. Number of comorbidities, median D-dimer and C-reactive protein increased with age. During COVID-19, AST/ALT and ALP/GGT levels significantly increased. Elevated hepatocellular transaminases (AST/ALT) and cholestasis parameters (ALP/GGT/BIL) were observed in 40.3% (n  = 262/650) and 45.0% (n  = 287/638) of patients respectively. Liver-related mortality was highest among patients with pre-existing decompensated liver disease (28.6%, p < .001). 1.7% of patients without pre-existing liver disease died of liver-related causes, that is consequences of hepatic dysfunction or acute liver failure. Importantly, COVID-19-associated liver injury (16.0%, p < .001), abnormal liver chemistries and liver-related mortality (6.5%, p < .001) were most frequent among 40-69 years old patients. Elevated AST and BIL after the first positive SARS-CoV-2 PCR independently predicted mortality in the overall cohort and in 40-69 years old patients.

CONCLUSIONS

Almost half of the COVID-19 patients exhibit abnormal hepatocellular and cholestasis-related liver chemistries with 40-69 years old patients being at particularly high risk for COVID-19-related liver injury and liver-related mortality. Elevated AST and BIL after SARS-CoV-2 infection are independent predictors of mortality, especially in patients aged 40-69 years.

A retrospective study of sugammadex for reversal of neuromuscular blockade induced by rocuronium in critically ill patients in the ICU

Sugammadex has been approved for reversal of neuromuscular blockade by vecuronium and rocuronium in adults undergoing surgery. Although widely used in the operating room, sugammadex has not been investigated in the intensive care unit setting. This study aimed to evaluate the use of sugammadex in critically ill patients with a focus on known drug-related adverse reactions. In this single-center, retrospective, observational study, 91 critically ill patients who were administered sugammadex while in the ICU were evaluated. Electronic health records were reviewed, and baseline data, as well as indication and incidence of complications possibly related to sugammadex, were retrospectively collected. The most common procedures requiring neuromuscular blockade followed by reversal with sugammadex were bronchoscopy, percutaneous dilatative tracheostomy, and percutaneous endoscopic gastrostomy. Within 2 h following administration of sugammadex, skin rash and use of antihistamines were reported in 4 patients (4.4%) in total; bradycardia was observed in 9 patients (9.9%), and respiratory adverse events were described in 3 patients (3.3%). New-onset bleeding up to 24 h after sugammadex was reported in 7 patients (7.7%), 3of whom received transfusions of packed red blood cells. Sugammadex was well tolerated in critically ill patients and could be considered for reversal of neuromuscular blockade in this population. Larger prospective studies are required to determine the safety profile and evaluate the potential benefit and indications of sugammadex in the critical care setting.

Duration of invasive mechanical ventilation prior to extracorporeal membrane oxygenation is not associated with survival in acute respiratory distress syndrome caused by coronavirus disease 2019

Background

Duration of invasive mechanical ventilation (IMV) prior to extracorporeal membrane oxygenation (ECMO) affects outcome in acute respiratory distress syndrome (ARDS). In coronavirus disease 2019 (COVID-19) related ARDS, the role of pre-ECMO IMV duration is unclear. This single-centre, retrospective study included critically ill adults treated with ECMO due to severe COVID-19-related ARDS between 01/2020 and 05/2021. The primary objective was to determine whether duration of IMV prior to ECMO cannulation influenced ICU mortality.

Results

During the study period, 101 patients (mean age 56 [SD ± 10] years; 70 [69%] men; median RESP score 2 [IQR 1–4]) were treated with ECMO for COVID-19. Sixty patients (59%) survived to ICU discharge. Median ICU length of stay was 31 [IQR 20.7–51] days, median ECMO duration was 16.4 [IQR 8.7–27.7] days, and median time from intubation to ECMO start was 7.7 [IQR 3.6–12.5] days. Fifty-three (52%) patients had a pre-ECMO IMV duration of > 7 days. Pre-ECMO IMV duration had no effect on survival (p = 0.95). No significant difference in survival was found when patients with a pre-ECMO IMV duration of < 7 days (< 10 days) were compared to ≥ 7 days (≥ 10 days) (p = 0.59 and p = 1.0).

Conclusions

The role of prolonged pre-ECMO IMV duration as a contraindication for ECMO in patients with COVID-19-related ARDS should be scrutinised. Evaluation for ECMO should be assessed on an individual and patient-centred basis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13613-022-00980-3.

Can We Protect Those We Care for in A Pandemic? - Prevalence of Neutralizing Antibodies against SARS-CoV-2 in Nursing Homes

In December 2019, the People's Republic of China and the World Health Organization first reported on a cluster of pneumonia with an unknown cause. Nine months later more than 1.4 million people have died from COVID 19. In this work, the effects of the COVID 19 pandemic on five nursing homes in Austria, which cared for 889 residents in the first half of 2020, were examined. The research question was whether the measures taken were appropriate to prevent an outbreak within the individual facilities. To detect previously unrecognized infections, the present study evaluated the prevalence of neutralizing antibodies against the SARS-CoV-2 virus in residents and employees of the nursing homes. Following the analysis of blood samples, the prospectively collected data was connected to data from screening examinations and data from contact tracing. The present study demonstrated an overall prevalence of neutralizing antibodies against the SARS-CoV-2 virus in nursing homes of 3.7%. Whereas the prevalence in those facilities that have never been hit by an outbreak is 0%, the prevalence in those facilities with an outbreak is up to 4.9%. Neutralizing antibodies against SARS-CoV-2 were detected in 35 persons. A retrospective analysis of all 5 included nursing homes demonstrated that upon regular clinical screening in combination with PCRs an infection with SARS-COV-2 was detected in 66 residents and 24 employees from different professional groups. In only 25 of the 35 persons with neutralizing antibodies against SARS-CoV-2 an infection was proven in advance. This study suggests that specific measures can prevent transmission within a health care facility. Nevertheless, the results also show that a risk reduction to 0% cannot be achieved. In preparation for further pandemic waves there is still the need to reduce the probability of a transmission in nursing homes with specific test strategies.

Virtual and Augmented Reality Applications in Medicine: Analysis of the Scientific Literature

BACKGROUND

Virtual reality (VR) and augmented reality (AR) have recently become popular research themes. However, there are no published bibliometric reports that have analyzed the corresponding scientific literature in relation to the application of these technologies in medicine.

OBJECTIVE

We used a bibliometric approach to identify and analyze the scientific literature on VR and AR research in medicine, revealing the popular research topics, key authors, scientific institutions, countries, and journals. We further aimed to capture and describe the themes and medical conditions most commonly investigated by VR and AR research.

METHODS

The Web of Science electronic database was searched to identify relevant papers on VR research in medicine. Basic publication and citation data were acquired using the "Analyze" and "Create Citation Report" functions of the database. Complete bibliographic data were exported to VOSviewer and Bibliometrix, dedicated bibliometric software packages, for further analyses. Visualization maps were generated to illustrate the recurring keywords and words mentioned in the titles and abstracts.

RESULTS

The analysis was based on data from 8399 papers. Major research themes were diagnostic and surgical procedures, as well as rehabilitation. Commonly studied medical conditions were pain, stroke, anxiety, depression, fear, cancer, and neurodegenerative disorders. Overall, contributions to the literature were globally distributed with heaviest contributions from the United States and United Kingdom. Studies from more clinically related research areas such as surgery, psychology, neurosciences, and rehabilitation had higher average numbers of citations than studies from computer sciences and engineering.

CONCLUSIONS

The conducted bibliometric analysis unequivocally reveals the versatile emerging applications of VR and AR in medicine. With the further maturation of the technology and improved accessibility in countries where VR and AR research is strong, we expect it to have a marked impact on clinical practice and in the life of patients.

Outcomes of non-COVID-19 critically ill patients during the COVID-19 pandemic : A retrospective propensity score-matched analysis

BACKGROUND

Coronavirus disease 2019 (COVID-19) disrupts routine care and alters treatment pathways in every medical specialty, including intensive care medicine, which has been at the core of the pandemic response. The impact of the pandemic is inevitably not limited to patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and their outcomes; however, the impact of COVID-19 on intensive care has not yet been analyzed.

METHODS

The objective of this propensity score-matched study was to compare the clinical outcomes of non-COVID-19 critically ill patients with the outcomes of prepandemic patients. Critically ill, non-COVID-19 patients admitted to the intensive care unit (ICU) during the first wave of the pandemic were matched with patients admitted in the previous year. Mortality, length of stay, and rate of readmission were compared between the two groups after matching.

RESULTS

A total of 211 critically ill SARS-CoV‑2 negative patients admitted between 13 March 2020 and 16 May 2020 were matched to 211 controls, selected from a matching pool of 1421 eligible patients admitted to the ICU in 2019. After matching, the outcomes were not significantly different between the two groups: ICU mortality was 5.2% in 2019 and 8.5% in 2020, p = 0.248, while intrahospital mortality was 10.9% in 2019 and 14.2% in 2020, p = 0.378. The median ICU length of stay was similar in 2019: 4 days (IQR 2-6) compared to 2020: 4 days (IQR 2-7), p = 0.196. The rate of ICU readmission was 15.6% in 2019 and 10.9% in 2020, p = 0.344.

CONCLUSION

In this retrospective single center study, mortality, ICU length of stay, and rate of ICU readmission did not differ significantly between patients admitted to the ICU during the implementation of hospital-wide COVID-19 contingency planning and patients admitted to the ICU before the pandemic.