Disease trajectories in interstitial lung diseases - data from the EXCITING-ILD registry.

BACKGROUND: Interstitial lung diseases (ILD) comprise a heterogeneous group of mainly chronic lung diseases with different disease trajectories. Progression (PF-ILD) occurs in up to 50% of patients and is associated with increased mortality. METHODS: The EXCITING-ILD (Exploring Clinical and Epidemiological Characteristics of Interstitial Lung Diseases) registry was analysed for disease trajectories in different ILD. The course of disease was classified as significant (absolute forced vital capacity FVC decline > 10%) or moderate progression (FVC decline 5-10%), stable disease (FVC decline or increase < 5%) or improvement (FVC increase ≥ 5%) during time in registry. A second definition for PF-ILD included absolute decline in FVC % predicted ≥ 10% within 24 months or ≥ 1 respiratory-related hospitalisation. Risk factors for progression were determined by Cox proportional-hazard models and by logistic regression with forward selection. Kaplan-Meier curves were utilised to estimate survival time and time to progression. RESULTS: Within the EXCITING-ILD registry 28.5% of the patients died (n = 171), mainly due to ILD (n = 71, 41.5%). Median survival time from date of diagnosis on was 15.5 years (range 0.1 to 34.4 years). From 601 included patients, progression was detected in 50.6% of the patients (n = 304) with shortest median time to progression in idiopathic NSIP (iNSIP; median 14.6 months) and idiopathic pulmonary fibrosis (IPF; median 18.9 months). Reasons for the determination as PF-ILD were mainly deterioration in lung function (PFT; 57.8%) and respiratory hospitalisations (40.6%). In multivariate analyses reduced baseline FVC together with age were significant predictors for progression (OR = 1.00, p < 0.001). Higher GAP indices were a significant risk factor for a shorter survival time (GAP stage III vs. I HR = 9.06, p < 0.001). A significant shorter survival time was found in IPF compared to sarcoidosis (HR = 0.04, p < 0.001), CTD-ILD (HR = 0.33, p < 0.001), and HP (HR = 0.30, p < 0.001). Patients with at least one reported ILD exacerbation as a reason for hospitalisation had a median survival time of 7.3 years (range 0.1 to 34.4 years) compared to 19.6 years (range 0.3 to 19.6 years) in patients without exacerbations (HR = 0.39, p < 0.001). CONCLUSION: Disease progression is common in all ILD and associated with increased mortality. Most important risk factors for progression are impaired baseline forced vital capacity and higher age, as well as acute exacerbations and respiratory hospitalisations for mortality. Early detection of progression remains challenging, further clinical criteria in addition to PFT might be helpful.

Pt-Black-Modified (Hemi)spherical AFM Sensors: In Situ Imaging of Light-Driven Hydrogen Peroxide Evolution.

In this work, we present (hemi)spherical atomic force microscopy (AFM) sensors for the detection of hydrogen peroxide. Platinum-black (Pt-B) was electrodeposited onto conductive colloidal AFM probes or directly at recessed microelectrodes located at the end of a tipless cantilever, resulting in electrocatalytically active cantilever-based sensors that have a small geometric area but, due to the porosity of the films, exhibit a large electroactive surface area. Focused ion beam-scanning electron microscopy tomography revealed the porous 3D structure of the deposited Pt-B. Given the accurate positioning capability of AFM, these probes are suitable for local in situ sensing of hydrogen peroxide and at the same time can be used for (electrochemical) force spectroscopy measurements. Detection limits for hydrogen peroxide in the nanomolar range (LOD = 68 ± 7 nM) were obtained. Stability test and first in situ proof-of-principle experiments to achieve the electrochemical imaging of hydrogen peroxide generated at a microelectrode and at photocatalytically active structured poly(heptazine imide) films are demonstrated. Force spectroscopic data of the photocatalyst films were recorded in ambient conditions, in solution, and by applying a potential, which demonstrates the versatility of these novel Pt-B-modified spherical AFM probes.

Hospitalisation patterns in interstitial lung diseases: data from the EXCITING-ILD registry.

BACKGROUND: Interstitial lung diseases (ILD) comprise a heterogeneous group of mainly chronic lung diseases with more than 200 entities and relevant differences in disease course and prognosis. Little data is available on hospitalisation patterns in ILD. METHODS: The EXCITING-ILD (Exploring Clinical and Epidemiological Characteristics of Interstitial Lung Diseases) registry was analysed for hospitalisations. Reasons for hospitalisation were classified as all cause, ILD-related and respiratory hospitalisations, and patients were analysed for frequency of hospitalisations, time to first non-elective hospitalisation, mortality and progression-free survival. Additionally, the risk for hospitalisation according to GAP index and ILD subtype was calculated by Cox proportional-hazard models as well as influencing factors on prediction of hospitalisation by logistic regression with forward selection. RESULTS: In total, 601 patients were included. 1210 hospitalisations were recorded during the 6 months prior to registry inclusion until the last study visit. 800 (66.1%) were ILD-related, 59.3% of admissions were registered in the first year after inclusion. Mortality was associated with all cause, ILD-related and respiratory-related hospitalisation. Risk factors for hospitalisation were advanced disease (GAP Index stages II and III) and CTD (connective tissue disease)-ILDs. All cause hospitalisations were associated with pulmonary hypertension (OR 2.53, p = 0.005). ILD-related hospitalisations were associated with unclassifiable ILD and concomitant emphysema (OR = 2.133, p = 0.001) as well as with other granulomatous ILDs and a positive smoking status (OR = 3.082, p = 0.005). CONCLUSION: Our results represent a crucial contribution in understanding predisposing factors for hospitalisation in ILD and its major impact on mortality. Further studies to characterize the most vulnerable patient group as well as approaches to prevent hospitalisations are warranted.

Facial disgust in response to touches, smells, and tastes.

Disgust serves to defend the body from the entry of toxins and disease. Central to this function is a strong relationship with the proximate senses of smell, taste, and touch. Theory suggests that distinct and reflexive facial movements should be evoked by gustatory and olfactory disgusts, serving to impede bodily entry. While this hypothesis has received some support from facial recognition studies, whether smell and taste disgusts actually produce distinct facial responses, is unknown. Moreover, there has been no assessment of the facial response evoked by contact with disgusting objects. To address these issues, this study compared facial responses to touch, smell, and taste disgusts. Sixty-four participants were asked to touch, smell, and taste disgust-evoking and neutral control stimuli, and rate them on disgust, on two occasions-first, while they were video recorded and second, with facial electromyography (EMG) applied (measuring levator labii and corrugator supercilii activity). Videos were coded for facial expressions by humans and for facial action units (FAUs) by machines. Self-report data confirmed the disgust stimuli as highly disgusting. Comparison of the overall pattern of FAUs evoked by touch, smell, and taste disgusts, indicated two distinct facial disgusts for the proximate senses-a chemosensory and a tactile-disgust face. The nose wrinkle and upper lip raise were central to all facial disgusts, indicating their centrality to the disgust face. Several facial disgusts appear to exist, each with different functional goals. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

[The virtual ILD board: a pilot project]. / Das virtuelle ILD-Board ­ Erfahrungen aus einem Pilotprojekt.

Interstitial lung diseases are associated with high morbitity and mortality. Rapid diagnosis in a qualified center is necessary in order to provide the best possible treatment. However, geographic distance and organizational issues lead to unacceptable delays. To support pulmonologists in private practice, we have trialed a digital system that minimizes such delays. The "virtual ILD board" leads to a considerably faster diagnosis and is a helpful tool for pulmonologists in practice. Standardization increases patient safety by ensuring interdisciplinary assessment and thus makes a relevant contribution to the management and guideline-based care of interstitial lung diseases.

The effect of sodium thiosulfate on immune cell metabolism during porcine hemorrhage and resuscitation.

Introduction: Sodium thiosulfate (Na2S2O3), an H2S releasing agent, was shown to be organ-protective in experimental hemorrhage. Systemic inflammation activates immune cells, which in turn show cell type-specific metabolic plasticity with modifications of mitochondrial respiratory activity. Since H2S can dose-dependently stimulate or inhibit mitochondrial respiration, we investigated the effect of Na2S2O3 on immune cell metabolism in a blinded, randomized, controlled, long-term, porcine model of hemorrhage and resuscitation. For this purpose, we developed a Bayesian sampling-based model for 13C isotope metabolic flux analysis (MFA) utilizing 1,2-13C2-labeled glucose, 13C6-labeled glucose, and 13C5-labeled glutamine tracers. Methods: After 3 h of hemorrhage, anesthetized and surgically instrumented swine underwent resuscitation up to a maximum of 68 h. At 2 h of shock, animals randomly received vehicle or Na2S2O3 (25 mg/kg/h for 2 h, thereafter 100 mg/kg/h until 24 h after shock). At three time points (prior to shock, 24 h post shock and 64 h post shock) peripheral blood mononuclear cells (PBMCs) and granulocytes were isolated from whole blood, and cells were investigated regarding mitochondrial oxygen consumption (high resolution respirometry), reactive oxygen species production (electron spin resonance) and fluxes within the metabolic network (stable isotope-based MFA). Results: PBMCs showed significantly higher mitochondrial O2 uptake and lower O 2 • - production in comparison to granulocytes. We found that in response to Na2S2O3 administration, PBMCs but not granulocytes had an increased mitochondrial oxygen consumption combined with a transient reduction of the citrate synthase flux and an increase of acetyl-CoA channeled into other compartments, e.g., for lipid biogenesis. Conclusion: In a porcine model of hemorrhage and resuscitation, Na2S2O3 administration led to increased mitochondrial oxygen consumption combined with stimulation of lipid biogenesis in PBMCs. In contrast, granulocytes remained unaffected. Granulocytes, on the other hand, remained unaffected. O 2 • - concentration in whole blood remained constant during shock and resuscitation, indicating a sufficient anti-oxidative capacity. Overall, our MFA model seems to be is a promising approach for investigating immunometabolism; especially when combined with complementary methods.

[Smoking Cessation in Patients with COPD]. / Tabakentwöhnung bei COPD.

Tobacco dependence is a common comorbidity in patients with COPD (Chronic Obstructive Pulmonary Disease) that negatively affects the course of the disease. However, clinically relevant improvement in COPD can only be achieved by complete and permanent abstinence. Therefore, abstinence from tobacco use is a central therapeutic concept in smoking patients with COPD and requires specific and targeted treatment.After detailed documentation of smoking behaviour and motivational counseling outlining the risks of smoking, all such patients shall be offered a structured therapy for tobacco cessation. There is high-quality evidence for the effectiveness of a combination therapy of behavioral therapy and medication (to treat the withdrawal syndrome). Due to insufficient data, there is currently no recommendation for the use of e-cigarettes as a primary option for a cessation attempt.Smoking is the most important cause of COPD. Smoking cessation is the most effective and cost-efficient single intervention to reduce the risk of developing and progressing COPD.

Mortality in cancer patients with SARS-CoV-2 or seasonal influenza: an observational cohort study from a German-wide hospital network.

PURPOSE: At the beginning of the COVID-19 pandemic, SARS-CoV-2 was often compared to seasonal influenza. We aimed to compare the outcome of hospitalized patients with cancer infected by SARS-CoV-2 or seasonal influenza including intensive care unit admission, mechanical ventilation and in-hospital mortality. METHODS: We analyzed claims data of patients with a lab-confirmed SARS-CoV-2 or seasonal influenza infection admitted to one of 85 hospitals of a German-wide hospital network between January 2016 and August 2021. RESULTS: 29,284 patients with COVID-19 and 7442 patients with seasonal influenza were included. Of these, 360 patients with seasonal influenza and 1625 patients with COVID-19 had any kind of cancer. Cancer patients with COVID-19 were more likely to be admitted to the intensive care unit than cancer patients with seasonal influenza (29.4% vs 24.7%; OR 1.31, 95% CI 1.00-1.73 p < .05). No statistical significance was observed in the mechanical ventilation rate for cancer patients with COVID-19 compared to those with seasonal influenza (17.2% vs 13.6% OR 1.34, 95% CI 0.96-1.86 p = .09). 34.9% of cancer patients with COVID-19 and 17.9% with seasonal influenza died (OR 2.45, 95% CI 1.81-3.32 p < .01). Risk factors among cancer patients with COVID-19 or seasonal influenza for in-hospital mortality included the male gender, age, a higher Elixhauser comorbidity index and metastatic cancer. CONCLUSION: Among cancer patients, SARS-CoV-2 was associated with a higher risk for in-hospital mortality than seasonal influenza. These findings underline the need of protective measurements to prevent an infection with either COVID-19 or seasonal influenza, especially in this high-risk population.

[The Usefulness of ERAS Concepts for Colorectal Resections - an Economic Analysis under DRG Conditions]. / Sinnhaftigkeit von ERAS-Konzepten bei kolorektalen Resektionen ­ eine ökonomische Analyse unter DRG-Bedingungen.

BACKGROUND: ERAS (Enhanced Recovery After Surgery) describes a multimodal, interdisciplinary and interprofessional treatment concept that optimizes the postoperative convalescence of the patient through the use of evidence-based measures. GOAL OF THE WORK: The aim of this article is to examine the economic feasibility of the concept in the German DRG system. MATERIAL AND METHODS: Since August 2019, patients have been treated in our clinic according to the later certified ERAS concept. The last 20 patients before ERAS implementation are compared below with 20 patients after ERAS implementation, who were identified using a matched pair analysis. In addition to the comparison of costs and revenues, the clinical outcome of the patients is also presented. RESULTS: The cases of the patients in the pre-ERAS cohort caused median costs of € 7432.83. BWR of 3.38 were billable. The resulting DRG revenue for the patients in this group amounted to € 11325.78. The proceeds generated in the end amounted to € 4575.14. The cases of patients in the ERAS cohort resulted in costs of € 5582.96. BWR of 2.84 could be billed. The DRG proceeds for the patients in this group therefore amounted to € 10014.18. The profit generated was thus € 4993.84. DISCUSSION: The cost reduction generated by ERAS was comparable to the "loss" caused by the BWR decrease. ERAS is therefore also possible to cover costs in the German DRG system.

Machine learning-derived prediction of in-hospital mortality in patients with severe acute respiratory infection: analysis of claims data from the German-wide Helios hospital network.

BACKGROUND: Severe acute respiratory infections (SARI) are the most common infectious causes of death. Previous work regarding mortality prediction models for SARI using machine learning (ML) algorithms that can be useful for both individual risk stratification and quality of care assessment is scarce. We aimed to develop reliable models for mortality prediction in SARI patients utilizing ML algorithms and compare its performances with a classic regression analysis approach. METHODS: Administrative data (dataset randomly split 75%/25% for model training/testing) from years 2016-2019 of 86 German Helios hospitals was retrospectively analyzed. Inpatient SARI cases were defined by ICD-codes J09-J22. Three ML algorithms were evaluated and its performance compared to generalized linear models (GLM) by computing receiver operating characteristic area under the curve (AUC) and area under the precision-recall curve (AUPRC). RESULTS: The dataset contained 241,988 inpatient SARI cases (75 years or older: 49%; male 56.2%). In-hospital mortality was 11.6%. AUC and AUPRC in the testing dataset were 0.83 and 0.372 for GLM, 0.831 and 0.384 for random forest (RF), 0.834 and 0.382 for single layer neural network (NNET) and 0.834 and 0.389 for extreme gradient boosting (XGBoost). Statistical comparison of ROC AUCs revealed a better performance of NNET and XGBoost as compared to GLM. CONCLUSION: ML algorithms for predicting in-hospital mortality were trained and tested on a large real-world administrative dataset of SARI patients and showed good discriminatory performances. Broad application of our models in clinical routine practice can contribute to patients' risk assessment and quality management.

The usefulness of Enhanced Recovery After Surgery concepts for colorectal resections: an economic analysis under DRG conditions.

PURPOSE: ERAS® (Enhanced Recovery After Surgery) describes a multimodal, interdisciplinary, and interprofessional treatment concept that optimizes the postoperative convalescence of the patient through the use of evidence-based measures. Goal of the work. The aim of this article is to examine the economic feasibility of the ERAS® concept in the German DRG (diagnosis-related groups) system. MATERIAL AND METHODS: Since August 2019, patients have been treated in our clinic according to the later certified ERAS® concept. The last 50 patients before ERAS® implementation are compared below with 50 patients after ERAS® implementation, who were identified using a matched pair analysis. In addition to the comparison of costs and revenues, the clinical outcome of the patients is also presented. RESULTS: The cases of the patients in the pre-ERAS® cohort caused median costs of € 7432.83. BWR (valuation ratio) of 3.38 were billable. The resulting DRG revenue for the patients in this group amounted to € 11,325.78. The proceeds generated in the end amounted to € 4575.14. The cases of patients in the ERAS® cohort resulted in costs of € 5582.96. BWR of 2.84 could be billed. The DRG proceeds for the patients in this group therefore amounted to € 10,014.18. The profit generated was thus € 4993.84. CONCLUSION: The cost reduction generated by ERAS® was more pronounced than the "loss" due to the decrease in BWR. ERAS® is therefore also possible in the German DRG system at absolutely cost-covering levels.

Changes in nationwide in-hospital stroke care during the first four waves of COVID-19 in Germany.

Introduction: In the early stages of the global COVID-19 pandemic hospital admissions for acute ischemic stroke (AIS) decreased substantially. As health systems have become more experienced in dealing with the pandemic, and as the proportion of the population vaccinated rises, it is of interest to determine whether the prevalence of AIS hospitalization and outcomes from hospitalization have returned to normal. Patients and methods: In this observational, retrospective cohort study, we compared the prevalence and outcomes of AIS during the first four waves of the pandemic to corresponding pre-pandemic periods in 2019 using administrative data collected from a nationwide network of 76 hospitals that manages 7% of all in-hospital cases in Germany. Results: We included 25,821 AIS cases in the study period (2020/2021) and used 26,295 AIS cases as controls (2019). Compared to pre-pandemic numbers, mean daily AIS admissions decreased only during wave 1 (from 39.6 to 34.1; p < 0.01) and wave 2 (from 39.9 to 38.3; p = 0.03) and returned to normal levels during waves 3 and 4. AIS case fatality increased in wave 1 only (from 6.0% to 7.6%; p = 0.03). We observed a consistent decrease in the prevalences of arterial hypertension, diabetes, and obesity among AIS cases throughout the pandemic and no changes in rates of systemic thrombolysis, mechanical thrombectomy, or decompressive craniectomy. The rate of transfer to stroke units increased only during waves 2 (by 4.6%; p < 0.01) and 3 (by 3.0%; p < 0.01). The proportion of patients with coinciding SARS-CoV-2 and AIS was low, peaking at 3.4% in wave 2 and subsequently decreasing to 0.4% in wave 4. Conclusion: In Germany, the COVID-19 pandemic seems to have had a larger effect on nationwide in-hospital AIS care during the early pandemic stages, in which AIS case numbers decreased and case fatality rose. This may reflect a nationwide "learning curve" within health care systems in providing AIS care in times of a pandemic.

Effect of gender, age and vaccine on reactogenicity and incapacity to work after COVID-19 vaccination: a survey among health care workers.

BACKGROUND: The aim of our study was to assess the impact the impact of gender and age on reactogenicity to three COVID-19 vaccine products: Biontech/Pfizer (BNT162b2), Moderna (mRNA-1273) and AstraZeneca (ChAdOx). Additional analyses focused on the reduction in working capacity after vaccination and the influence of the time of day when vaccines were administered. METHODS: We conducted a survey on COVID-19 vaccinations and eventual reactions among 73,000 employees of 89 hospitals of the Helios Group. On May 19th, 2021 all employees received an email, inviting all employees who received at least 1 dose of a COVID-19 to participate using an attached link. Additionally, the invitation was posted in the group's intranet page. Participation was voluntary and non-traceable. The survey was closed on June 21st, 2021. RESULTS: 8375 participants reported on 16,727 vaccinations. Reactogenicity was reported after 74.6% of COVID-19 vaccinations. After 23.0% vaccinations the capacity to work was affected. ChAdOx induced impairing reactogenicity mainly after the prime vaccination (70.5%), while mRNA-1273 led to more pronounced reactions after the second dose (71.6%). Heterologous prime-booster vaccinations with ChAdOx followed by either mRNA-1273 or BNT162b2 were associated with the highest risk for impairment (81.4%). Multivariable analyses identified the factors older age, male gender and vaccine BNT162b as independently associated with lower odds ratio for both, impairing reactogenicity and incapacity to work. In the comparison of vaccine schedules, the heterologous combination ChAdOx + BNT162b or mRNA-1273 was associated with the highest and the homologue prime-booster vaccination with BNT162b with the lowest odds ratios. The time of vaccination had no significant influence. CONCLUSIONS: Around 75% of the COVID-19 vaccinations led to reactogenicity and nearly 25% of them led to one or more days of work loss. Major risk factors were female gender, younger age and the administration of a vaccine other than BNT162b2. When vaccinating a large part of a workforce against COVID-19, especially in professions with a higher proportion of young and women such as health care, employers and employees must be prepared for a noticeable amount of absenteeism. Assuming vaccine effectiveness to be equivalent across the vaccine combinations, to minimize reactogenicity, employees at risk should receive a homologous prime-booster immunisation with BNT162b2. TRIAL REGISTRATION: The study was approved by the Ethic Committee of the Aerztekammer Berlin on May 27th, 2021 (Eth-37/21) and registered in the German Clinical Trials Register (DRKS 00025745). The study was supported by the Helios research grant HCRI-ID 2021-0272.

A Comparative Analysis of In-Hospital Mortality per Disease Groups in Germany Before and During the COVID-19 Pandemic From 2016 to 2020.

Importance: Throughout the ongoing SARS-CoV-2 pandemic, it has been critical to understand not only the viral disease itself but also its implications for the overall health care system. Reports about excess mortality in this regard have mostly focused on overall death counts during specific pandemic phases. Objective: To investigate hospitalization rates and compare in-hospital mortality rates with absolute mortality incidences across a broad spectrum of diseases, comparing 2020 data with those of prepandemic years. Design, Setting, and Participants: Retrospective, cross-sectional, multicentric analysis of administrative data from 5 821 757 inpatients admitted from January 1, 2016, to December 31, 2020, to 87 German Helios primary to tertiary care hospitals. Exposures: Exposure to SARS-CoV-2. Main Outcomes and Measures: Administrative data were analyzed from January 1, 2016, to March 31, 2021, as a consecutive sample for all inpatients. Disease groups were defined according to International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10; German modification) encoded main discharge diagnoses. Incidence rate ratios (IRRs) for hospital admissions and hospital mortality counts, as well as relative mortality risks (RMRs) comparing 2016-2019 with 2020 (exposure to the SARS-CoV-2 pandemic), were calculated with Poisson regression with log-link function. Results: Data were examined for 5 821 757 inpatients (mean [SD] age, 56.4 [25.3] years; 51.5% women), including 125 807 in-hospital deaths. Incidence rate ratios for hospital admissions were associated with a significant reduction for all investigated disease groups (IRR, 0.82; 95% CI, 0.79-0.86; P < .001). After adjusting for age, sex, the Elixhauser Comorbidity Index score, and SARS-CoV-2 infections, RMRs were associated with an increase in infectious diseases (RMR, 1.28; 95% CI, 1.21-1.34; P < .001), musculoskeletal diseases (RMR, 1.19; 95% CI, 1.04-1.36; P = .009), and respiratory diseases (RMR, 1.09; 95% CI, 1.05-1.14; P < .001) but not for the total cohort (RMR, 1.00; 95% CI, 0.99-1.02; P = .66). Regarding in-hospital mortality, IRR was associated with an increase within the ICD-10 chapter of respiratory diseases (IRR, 1.28; 95% CI, 1.13-1.46; P < .001) in comparing 2020 with 2016-2019, in contrast to being associated with a reduction in IRRs for the overall cohort and several other subgroups. After exclusion of patients with SARS-CoV-2 infections, IRRs were associated with a reduction in absolute in-hospital mortality for the overall cohort (IRR, 0.78; 95% CI, 0.72-0.84; P < .001) and the subgroup of respiratory diseases (IRR, 0.83; 95% CI, 0.74-0.92; P < .001). Conclusions and Relevance: This cross-sectional study of inpatients from a multicentric German database suggests that absolute in-hospital mortality for 2020 across disease groups was not higher compared with previous years. Higher IRRs of in-hospital deaths observed in patients with respiratory diseases were likely associated with individuals with SARS-CoV-2 infections.

[Psychiatric Emergency Admissions and Inpatient Length of Stay Before and During the COVID-19 Pandemic in Germany]. / Psychiatrische Notfallaufnahmen und stationäre Aufenthaltsdauer vor und während der COVID-19-Pandemie in Deutschland.

OBJECTIVE: The impact of the COVID-19 year on the number of daily psychiatric emergency admissions and length of stay was compared with previous years. METHODS: In a retrospective study, the four quarters of 2020 of several psychiatric hospitals in Germany were statistically compared with the respective quarters of 2018 and 2019. RESULTS: A total of 73,412 cases was analyzed. In the 2nd quarter of 2020, the number of daily admissions was significantly lower as compared to the control period (59.1 vs. 70.7; incidence rate ratio [95 % confidence interval] 0.81 [0.69; 0.96]; p = 0.012). Length of stay was numerically but not significantly different as compared to the control periods. CONCLUSION: The COVID-19 pandemic had a strong impact on inpatient psychiatric care. In the future, multimodal care structures must ensure the care of severely mentally ill people in crisis situations.

From population- to patient-based prediction of in-hospital mortality in heart failure using machine learning.

Aims: Utilizing administrative data may facilitate risk prediction in heart failure inpatients. In this short report, we present different machine learning models that predict in-hospital mortality on an individual basis utilizing this widely available data source. Methods and results: Inpatient cases with a main discharge diagnosis of heart failure hospitalized between 1 January 2016 and 31 December 2018 in one of 86 German Helios hospitals were examined. Comorbidities were defined by ICD-10 codes from administrative data. The data set was randomly split into 75/25% portions for model development and testing. Five algorithms were evaluated: logistic regression [generalized linear models (GLMs)], random forest (RF), gradient boosting machine (GBM), single-layer neural network (NNET), and extreme gradient boosting (XGBoost). After model tuning, the receiver operating characteristics area under the curves (ROC AUCs) were calculated and compared with DeLong's test. A total of 59 074 inpatient cases (mean age 77.6 ± 11.1 years, 51.9% female, 89.4% NYHA Class III/IV) were included and in-hospital mortality was 6.2%. In the test data set, calculated ROC AUCs were 0.853 [95% confidence interval (CI) 0.842-0.863] for GLM, 0.851 (95% CI 0.840-0.862) for RF, 0.855 (95% CI 0.844-0.865) for GBM, 0.836 (95% CI 0.823-0.849) for NNET, and 0.856 (95% CI 9.846-0.867) for XGBoost. XGBoost outperformed all models except GBM. Conclusion: Machine learning-based processing of administrative data enables the creation of well-performing prediction models for in-hospital mortality in heart failure patients.

Expected and observed in-hospital mortality in heart failure patients before and during the COVID-19 pandemic: Introduction of the machine learning-based standardized mortality ratio at Helios hospitals.

BACKGROUND: Reduced hospital admission rates for heart failure (HF) and evidence of increased in-hospital mortality were reported during the COVID-19 pandemic. The aim of this study was to apply a machine learning (ML)-based mortality prediction model to examine whether the latter is attributable to differing case mixes and exceeds expected mortality rates. METHODS AND RESULTS: Inpatient cases with a primary discharge diagnosis of HF non-electively admitted to 86 German Helios hospitals between 01/01/2016 and 08/31/2020 were identified. Patients with proven or suspected SARS-CoV-2 infection were excluded. ML-based models were developed, tuned, and tested using cases of 2016-2018 (n = 64,440; randomly split 75%/25%). Extreme gradient boosting showed the best model performance indicated by a receiver operating characteristic area under the curve of 0.882 (95% confidence interval [CI]: 0.872-0.893). The model was applied on data sets of 2019 and 2020 (n = 28,556 cases) and the hospital standardized mortality ratio (HSMR) was computed as the observed to expected death ratio. Observed mortality rates were 5.84% (2019) and 6.21% (2020), HSMRs based on an individual case-based mortality probability were 100.0 (95% CI: 93.3-107.2; p = 1.000) for 2019 and 99.3 (95% CI: 92.5-106.4; p = .850) for 2020. Within subgroups of age or hospital volume, there were no significant differences between observed and expected deaths. When stratified for pandemic phases, no excess death during the COVID-19 pandemic was observed. CONCLUSION: Applying an ML algorithm to calculate expected inpatient mortality based on administrative data, there was no excess death above expected event rates in HF patients during the COVID-19 pandemic.