Due to the impact of COVID-19, a significant influx of emergency patients inundated the intensive care unit (ICU), and as a result, the treatment of elective patients was postponed or even cancelled. ...This paper studies ICU bed allocation for three categories of patients (emergency, elective, and current ICU patients). A two-stage model and an improved Non-dominated Sorting Genetic Algorithm II (NSGA-II) are used to obtain ICU bed allocation. In the first stage, bed allocation is examined under uncertainties regarding the number of emergency patients and their length of stay (LOS). In the second stage, in addition to including the emergency patients with uncertainties in the first stage, it also considers uncertainty in the LOS of elective and current ICU patients. The two-stage model aims to minimize the number of required ICU beds and maximize resource utilization while ensuring the admission of the maximum number of patients. To evaluate the effectiveness of the model and algorithm, the improved NSGA-II was compared with two other methods: multi-objective simulated annealing (MOSA) and multi-objective Tabu search (MOTS). Drawing on data from real cases at a hospital in Lyon, France, the NSGA-II, while catering to patient requirements, saves 9.8% and 5.1% of ICU beds compared to MOSA and MOTS. In five different scenarios, comparing these two algorithms, NSGA-II achieved average improvements of 0%, 49%, 11.4%, 9.5%, and 17.1% across the five objectives.
The Hospital Frailty Risk Score (HFRS) can be applied to medico-administrative datasets to determine the risks of 30-day mortality and long length of stay (LOS) in hospitalized older patients. The ...objective of this study was to compare the HFRS with Charlson and Elixhauser comorbidity indices, used separately or combined.
A retrospective analysis of the French medical information database. The HFRS, Charlson index, and Elixhauser index were calculated for each patient based on the index stay and hospitalizations over the preceding 2 years. Different constructions of the HFRS were considered based on overlapping diagnostic codes with either Charlson or Elixhauser indices. We used mixed logistic regression models to investigate the association between outcomes, different constructions of HFRS, and associations with comorbidity indices.
743 hospitals in France.
All patients aged 75 years or older hospitalized as an emergency in 2017 (n=1,042,234).Main outcome measures: 30-day inpatient mortality and LOS >10 days.
The HFRS, Charlson, and Elixhauser indices were comparably associated with an increased risk of 30-day inpatient mortality and long LOS. The combined model with the highest c-statistic was obtained when associating the HFRS with standard adjustment and Charlson for 30-day inpatient mortality (adjusted c-statistics: HFRS=0.654; HFRS + Charlson = 0.676) and with Elixhauser for long LOS (adjusted c-statistics: HFRS= 0.672; HFRS + Elixhauser =0.698).
Combining comorbidity indices and HFRS may improve discrimination for predicting long LOS in hospitalized older people, but adds little to Charlson's 30-day inpatient mortality risk.
Length of stay (LOS) is an important metric for the organization and scheduling of care activities. This study sought to propose a LOS prediction method based on deep learning using widely available ...administrative data from acute and emergency care and compare it with other methods.
All admissions between January 1, 2011 and December 31, 2019, at 6 university hospitals of the Hospices Civils de Lyon metropolis were included, leading to a cohort of 1,140,100 stays of 515,199 patients. Data included demographics, primary and associated diagnoses, medical procedures, the medical unit, the admission type, socio-economic factors, and temporal information. A model based on embeddings and a Feed-Forward Neural Network (FFNN) was developed to provide fine-grained LOS predictions per hospitalization step. Performances were compared with random forest and logistic regression, with the accuracy, Cohen kappa, and a Bland-Altman plot, through a 5-fold cross-validation.
The FFNN achieved an accuracy of 0.944 (CI: 0.937, 0.950) and a kappa of 0.943 (CI: 0.935, 0.950). For the same metrics, random forest yielded 0.574 (CI: 0.573, 0.575) and 0.602 (CI: 0.601, 0.603), respectively, and 0.352 (CI: 0.346, 0.358) and 0.414 (CI: 0.408, 0.422) for the logistic regression. The FFNN had a limit of agreement ranging from -2.73 to 2.67, which was better than random forest (-6.72 to 6.83) or logistic regression (-7.60 to 9.20).
The FFNN was better at predicting LOS than random forest or logistic regression. Implementing the FFNN model for routine acute care could be useful for improving the quality of patients' care.
The Hospital Frailty Risk Score (HFRS) has made it possible internationally to identify subgroups of patients with characteristics of frailty from routinely collected hospital data.
To externally ...validate the HFRS in France.
A retrospective analysis of the French medical information database.
743 hospitals in Metropolitan France.
All patients aged 75 years or older hospitalised as an emergency in 2017 (n = 1,042,234).
The HFRS was calculated for each patient based on the index stay and hospitalisations over the preceding 2 years. Main outcome measures were 30-day in-patient mortality, length of stay (LOS) >10 days and 30-day readmissions. Mixed logistic regression models were used to investigate the association between outcomes and HFRS score.
Patients with high HFRS risk were associated with increased risk of mortality and prolonged LOS (adjusted odds ratio aOR = 1.38 1.35-1.42 and 3.27 3.22-3.32, c-statistics = 0.676 and 0.684, respectively), while it appeared less predictive of readmissions (aOR = 1.00 0.98-1.02, c-statistic = 0.600). Model calibration was excellent. Restricting the score to data prior to index admission reduced discrimination of HFRS substantially.
HFRS can be used in France to determine risks of 30-day in-patient mortality and prolonged LOS, but not 30-day readmissions. Trial registration: Reference ID on clinicaltrials.gov: ID: NCT03905629.
Background
The COVID-19 sanitary crisis inflicted different challenges regarding the reorganization of the human and logistic resources, particularly in intensive care unit (ICU). Interdependence ...between regional pandemic burden and individual outcome remains unknown. The study aimed to assess the association between ICU bed occupancy and case fatality rate of critically ill COVID-19 patients.
Methods
A cross-sectional study was performed in France, using the national hospital discharge database from March to May, 2020. All patients admitted to ICU for COVID-19 were included. Case fatality was described according to: (i) patient’s characteristics (age, sex, comorbid conditions, ICU interventions); (ii) hospital’s characteristics (baseline ICU experience assessed by the number of ICU stays in 2019, number of ICU physicians per bed), and (iii) the regional outbreak-related profiles (workload indicator based on ICU bed occupancy). The determinants of lethal outcome were identified using a logistic regression model.
Results
14,513 COVID-19 patients were admitted to ICU; 4256 died (29.3%), with important regional inequalities in case fatality (from 17.6 to 33.5%). Older age, multimorbidity and clinical severity were associated with higher mortality, as well as a lower baseline ICU experience of the health structure. Regions with more than 10 days with ≥ 75% of ICU occupancy by COVID-19 patients experienced an excess of mortality (up to adjusted OR = 2.2 1.9–2.6 for region with the highest occupancy rate of ICU beds).
Conclusions
The regions with the highest burden of care in ICU were associated with up to 2.2-fold increase of death rate.
This study aimed to assess the performance improvement for machine learning-based hospital length of stay (LOS) predictions when clinical signs written in text are accounted for and compared to the ...traditional approach of solely considering structured information such as age, gender and major ICD diagnosis.
This study was an observational retrospective cohort study and analyzed patient stays admitted between 1 January to 24 September 2019. For each stay, a patient was admitted through the Emergency Department (ED) and stayed for more than two days in the subsequent service. LOS was predicted using two random forest models. The first included unstructured text extracted from electronic health records (EHRs). A word-embedding algorithm based on UMLS terminology with exact matching restricted to patient-centric affirmation sentences was used to assess the EHR data. The second model was primarily based on structured data in the form of diagnoses coded from the International Classification of Disease 10th Edition (ICD-10) and triage codes (CCMU/GEMSA classifications). Variables common to both models were: age, gender, zip/postal code, LOS in the ED, recent visit flag, assigned patient ward after the ED stay and short-term ED activity. Models were trained on 80% of data and performance was evaluated by accuracy on the remaining 20% test data.
The model using unstructured data had a 75.0% accuracy compared to 74.1% for the model containing structured data. The two models produced a similar prediction in 86.6% of cases. In a secondary analysis restricted to intensive care patients, the accuracy of both models was also similar (76.3% vs 75.0%).
LOS prediction using unstructured data had similar accuracy to using structured data and can be considered of use to accurately model LOS.
Abstract Background A previous study reported significant excess mortality among non-COVID-19 patients due to disrupted surgical care caused by resource prioritization for COVID-19 cases in France. ...The primary objective was to investigate if a similar impact occurred for medical conditions and determine the effect of hospital saturation on non-COVID-19 hospital mortality during the first year of the pandemic in France. Methods We conducted a nationwide population-based cohort study including all adult patients hospitalized for non-COVID-19 acute medical conditions in France between March 1, 2020 and 31 May, 2020 (1st wave) and September 1, 2020 and December 31, 2020 (2nd wave). Hospital saturation was categorized into four levels based on weekly bed occupancy for COVID-19: no saturation (< 5%), low saturation (> 5% and ≤ 15%), moderate saturation (> 15% and ≤ 30%), and high saturation (> 30%). Multivariate generalized linear model analyzed the association between hospital saturation and mortality with adjustment for age, sex, COVID-19 wave, Charlson Comorbidity Index, case-mix, source of hospital admission, ICU admission, category of hospital and region of residence. Results A total of 2,264,871 adult patients were hospitalized for acute medical conditions. In the multivariate analysis, the hospital mortality was significantly higher in low saturated hospitals (adjusted Odds Ratio/aOR = 1.05, 95% CI 1.34–1.07, P < .001), moderate saturated hospitals (aOR = 1.12, 95% CI 1.09–1.14, P < .001), and highly saturated hospitals (aOR = 1.25, 95% CI 1.21–1.30, P < .001) compared to non-saturated hospitals. The proportion of deaths outside ICU was higher in highly saturated hospitals (87%) compared to non-, low- or moderate saturated hospitals (81–84%). The negative impact of hospital saturation on mortality was more pronounced in patients older than 65 years, those with fewer comorbidities (Charlson 1–2 and 3 vs. 0), patients with cancer, nervous and mental diseases, those admitted from home or through the emergency room (compared to transfers from other hospital wards), and those not admitted to the intensive care unit. Conclusions Our study reveals a noteworthy “dose-effect” relationship: as hospital saturation intensifies, the non-COVID-19 hospital mortality risk also increases. These results raise concerns regarding hospitals’ resilience and patient safety, underscoring the importance of identifying targeted strategies to enhance resilience for the future, particularly for high-risk patients.
Abstract Background Immune-mediated haemolytic anaemia (IMHA) has a high mortality rate within the first weeks to months of diagnosis. Identifying dogs at increased risk of death may help guide ...decision-making for owners and veterinarians. Prior studies have identified several but inconsistent prognostic factors. The objectives of the study were to describe the clinical presentation and outcome of canine immune-mediated haemolytic anaemia in Ireland and to assess for independent factors associated with survival including long-term survival. Medical records from a single centre were reviewed between 2002 and 2020 to identify dogs with immune-mediated haemolytic anaemia using the American College of Veterinary Internal Medicine (ACVIM) consensus statement algorithm. Survival analysis was performed using univariable Cox proportional hazards regression models with Breslow method for ties to identify prognostic factors. Results One hundred and four cases were included. The diagnosis of immune-mediated haemolytic anaemia was classified as definitive, supportive and suspicious in 42 (40%), 50 (48%), and 12 dogs (12%) respectively. Twenty-two dogs (21%) were diagnosed with associative IMHA and 82 dogs were diagnosed with non-associative IMHA (79%). 65% of the cases received more than one immunosuppressive medication during the course of treatment. The mortality rate at one and three months was 16% (95% confidence interval CI 9–26) and 31% (95% CI 21–43) respectively. Excluding dogs that died within three months, the median survival time was 2664 days. The relapse rate during the follow-up period was 7%. Survival did not improve over the course of the study period. Thrombocytopenia and hyperbilirubinaemia were identified as negative prognostic indicators (Hazard ratio 2.2 and 2.5, 95% CI 1.1–4.1 and 1.1–5.6, respectively). Conclusions Excluding dogs that died within three months, the outcome was good in dogs with non-associative immune-mediated haemolytic anaemia in Ireland. The relapse rate was low regardless of the presence of associative causes. Thrombocytopenia and hyperbilirubinaemia were the only independent negative prognostic factors. The one-month and three-month mortality rates were similar compared to prior studies and survival did not improve over time during the study period: the mortality rate of canine immune-mediated haemolytic anaemia remains high in the acute phase.