Academic literature on the topic 'Mortality and Length of stay Prediction'

En médecine prédictive personnalisée, modéliser avec précision la maladie et les processus de soins d'un patient est crucial en raison des dépendances temporelles à long terme inhérentes. Cependant, les dossiers de santé électroniques (DSE) se composent souvent de données épisodiques et irrégulières, issues des admissions hospitalières sporadiques, créant des schémas uniques pour chaque séjour hospitalier.Par conséquent, la construction d'un modèle prédictif personnalisé nécessite une considération attentive de ces facteurs pour capturer avec précision le parcours de santé du patient et aider à la prise de décision clinique.LSTM sont efficaces pour traiter les données séquentielles comme les DSE, mais ils présentent deux limitations majeures : l'incapacité à interpréter les résultats des prédictions et à prendre en compte des intervalles de temps irréguliers entre les événements consécutifs. Pour surmonter ces limitations, nous introduisons de nouveaux réseaux neuronaux à mémoire dynamique profonde appelés Multi-Way Adaptive et Adaptive Multi-Way Interpretable Time-Aware LSTM (MWTA-LSTM etAMITA), conçus pour les données séquentielles collectées de manière irrégulière.L'objectif principal des deux modèles est de tirer parti des dossiers médicaux pour mémoriser les trajectoires de maladie et les processus de soins, estimer les états de maladie actuels et prédire les risques futurs, offrant ainsi un haut niveau de précision et de pouvoir prédictif<br>In personalized predictive medicine, accurately modeling a patient's illness and care processes is crucial due to the inherent long-term temporal dependencies. However, Electronic Health Records (EHRs) often consist of episodic and irregularly timed data, stemming from sporadic hospital admissions, which create unique patterns for each hospital stay. Consequently, constructing a personalized predictive model necessitates careful consideration of these factors to accurately capture the patient's health journey and assist in clinical decision-making. LSTM networks are effective for handling sequential data like EHRs, but they face two significant limitations: the inability to interpret prediction results and to take into account irregular time intervals between consecutive events. To address limitations, we introduce novel deep dynamic memory neural networks called Multi-Way Adaptive and Adaptive Multi-Way Interpretable Time-Aware LSTM (MWTA-LSTM and AMITA) designed for irregularly collected sequential data. The primary objective of both models is to leverage medical records to memorize illness trajectories and care processes, estimate current illness states, and predict future risks, thereby providing a high level of precision and predictive power

Class of 2009 Abstract<br>OBJECTIVES: The purpose of the study was to determine the rate of Clostridium difficile infections (CDI) in hospitalized patients and the various factors that were associated with the risk of developing CDI by examining patient discharge data for hospitals in 37 states in the United States using Healthcare Cost and Utilization Project (HCUP). METHODS: Patient discharge information for all patients obtained using HCUP census for the years 2002-2005, either for primary or secondary (all-listed) occurrences of CDI using the ICD-9-CM code (008.45) specific for intestinal infections due to C. difficile, were included in the study. Regression analysis, either Generalized Linear Model log-link or power-link, or a logistic regression was employed to control for the multiple independent variables. RESULTS: The incidence rate for CDI was 9.4% for the years 2002-2005. Among the concomitant diagnoses and procedures, essential hypertension, volume depletion, congestive heart failure, urinary tract infection and venous catheterization were the top 5. The length of stay (LOS) for CDI was associated with being Black, Hispanic or Other race category, number of diagnoses and procedures, primary expected payer of Medicaid, private insurance and other (including worker’s compensation, CHAMPUS,CHAMPVA etc), and all groups classified based on median household income category for patient’s zip code. Predictors of CDI related to inpatient total hospital charges were being female, race (other than black), number of diagnoses and procedures, Death, LOS, patient location and with self-pay and no charge categories as primary expected payer. Predictors of higher CDI related inpatient hospital deaths were age, female sex, Hispanic race, number of diagnoses and procedures, LOS and having Medicaid, self-pay or other as primary expected payer. CONCLUSIONS: LOS, inpatient total hospital charges, and inpatient mortality were dependent on several patient and other characteristics.

Class of 2009<br>OBJECTIVES: To determine and characterize the relative impact of patient demographics on hospital inpatient charges, length of stay, and inpatient mortality in patients with ovarian cancer from 2002-2005. METHODS: A retrospective database analysis of AHRQ’s Health Care Cost and Utilization Project (HCUP) Nationwide Inpatient Sample databases was conducted spanning from January 1, 2002, to December 31, 2005.Data were collected regarding age, race, payer status, median household income, location of hospital (urban/rural), comorbidities, procedures, total charges, length of stay, and inpatient mortality. Multivariate and gamma regression methods were utilized to examine incremental risks associated with length of stay, total charges, and inpatient mortality, after controlling for all other variables. RESULTS: Overall, data from 246,012 hospital admissions were obtained. The average length of stay of patients was 6.58 days (SD = 7.22), the average number of diagnoses was 7.18 (SD = 3.36), the average number of procedures performed was 2.71 (SD = 2.66). A total of 14,485 (5.9%) patients died during hospitalization. The average total charge was $29,698 (SD = $42,951). The IRR was 0.886 (95%CI, -0.105 to -0.04) for patients who were Hispanic, and 1.089 (95%CI, 0.017–0.153) for patients who were Black compared to patients who were white. When compared to patients who lived in large, metropolitan areas, the IRR was 0.88 (95%CI, -0.146 to - 0.109) for patients located in smaller, metropolitan areas, and the IRR was 0.74 (95%CI, -0.335 to -0.268) for patients located in non- urban areas. CONCLUSIONS: Patient demographics were found to have associations, both directly and indirectly, with length o

Recent advancements in machine learning and the widespread adoption of electronic healthrecords have enabled breakthroughs for several predictive modelling tasks in health care. One such task that has seen considerable improvements brought by deep neural networks is length of stay (LOS) prediction, in which research has mainly focused on adult patients in the intensive care unit. This thesis uses multivariate time series extracted from the publicly available Medical Information Mart for Intensive Care III database to explore the potential of deep learning for classifying the remaining LOS of newborns in the neonatal intensive care unit (NICU) at each hour of the stay. To investigate this, this thesis describes experiments conducted with various deep learning models, including long short-term memory cells, gated recurrentunits, fully-convolutional networks and several composite networks. This work demonstrates that modelling the remaining LOS of newborns in the NICU as a multivariate time series classification problem naturally facilitates repeated predictions over time as the stay progresses and enables advanced deep learning models to outperform a multinomial logistic regression baseline trained on hand-crafted features. Moreover, it shows the importance of the newborn’s gestational age and binary masks indicating missing values as variables for predicting the remaining LOS.<br>Framstegen inom maskininlärning och det utbredda införandet av elektroniska hälsoregister har möjliggjort genombrott för flera prediktiva modelleringsuppgifter inom sjukvården. En sådan uppgift som har sett betydande förbättringar förknippade med djupa neurala nätverk är förutsägelsens av vistelsetid på sjukhus, men forskningen har främst inriktats på vuxna patienter i intensivvården. Den här avhandlingen använder multivariata tidsserier extraherade från den offentligt tillgängliga databasen Medical Information Mart for Intensive Care III för att undersöka potentialen för djup inlärning att klassificera återstående vistelsetid för nyfödda i den neonatala intensivvårdsavdelningen (neonatal-IVA) vid varje timme av vistelsen. Denna avhandling beskriver experiment genomförda med olika djupinlärningsmodeller, inklusive longshort-term memory, gated recurrent units, fully-convolutional networks och flera sammansatta nätverk. Detta arbete visar att modellering av återstående vistelsetid för nyfödda i neonatal-IVA som ett multivariat tidsserieklassificeringsproblem på ett naturligt sätt underlättar upprepade förutsägelser över tid och gör det möjligt för avancerade djupa inlärningsmodeller att överträffaen multinomial logistisk regressionsbaslinje tränad på handgjorda funktioner. Dessutom visar det vikten av den nyfödda graviditetsåldern och binära masker som indikerar saknade värden som variabler för att förutsäga den återstående vistelsetiden.

RESUMO - Introdução: O presente estudo pretende analisar o impacte na saúde e a implicação nos custos da Pneumonia adquirida durante o internamento hospitalar. Está comprovado que as infeções hospitalares constituem um problema de Saúde pública dos hospitais em todo o mundo. Metodologia: A população em estudo abrange 97 033 episódios de internamento, ocorridos em 10 hospitais, no ano de 2010. O trabalho compreende três fases: i) caracterização da população em estudo; ii) identificação das variáveis que influenciam os resultados em saúde; iii) estimação dos custos do internamento com Pneumonia. Resultados: Os episódios de internamento com Pneumonia ocorreram maioritariamente no sexo masculino (58.1%). A faixa etária com mais episódios foi a dos 80 aos 89 anos. A taxa de Prevalência foi de 4.16% e a taxa de Mortalidade foi de 34.56%. Os doentes com Pneumonia tiveram uma demora média superior em 13 dias em relação aos doentes sem Pneumonia para o mesmo conjunto de GDH. Pertencer ao sexo masculino e os episódios de internamento ocorridos em hospitais não universitários levam a um aumento da probabilidade de morrer. Por sua vez apresentar uma maior duração de internamento e um número superior de comorbilidades levam a uma diminuição deste risco. Os custos em excesso dos episódios de internamento devido à aquisição de Pneumonia como doença secundária foram de aproximadamente 18 milhões de euros. Conclusão: O trabalho foi elaborado tendo em vista a quantificação do fenómeno em Portugal, tanto em termos da carga da doença, como das implicações financeiras. Os valores encontrados são preocupantes, pelo que se torna necessário tomar medidas e introduzir práticas na atividade hospitalar que minimizem as infeções hospitalares em geral e da Pneumonia em particular. Por sua vez é expectável, face ao descrito na literatura internacional, que a introdução destas práticas melhor os resultados em saúde e o desempenho financeiro dos hospitais.<br>ABSTRACT - Introduction: The main goal of this study is to analyze the health and the costs due to acquired Pneumonia during hospital stay. There is evidence that hospital infections are a public health problem in hospitals worldwide. Methods: The population analyzed is 97,033 hospital admissions, occurred in 10 hospitals in the year 2010. The work comprises three phases: i) characterization of the population, ii) identification of variables that influence health outcomes, iii) estimating the costs of acquired Pneumonia. Results: Admissions with acquired Pneumonia are more frequent on males (58.1%). The most relevant age group was from 80 to 89 years. The prevalence rate was 4.16% and the in-hospital mortality rate was 34.56%. The patients with acquired Pneumonia had an increase of the length of stay circa 13 days compared with patients without acquired Pneumonia for the same set of GDH. The males and admissions on non-teaching hospitals lead to an increased risk of hospital death. Moreover larger length of stay and higher number of comorbidities had decreased the risk of hospital death. The increase on admissions costs due to acquired Pneumonia were circa 18 million euros. Conclusions: The study presents some poor health outcomes, as well as costs increase due to acquired Pneumonia in Portuguese public hospitals. These results should be considered as a real problem in Portugal, and therefore it is necessary to be more evidenced based on hospital guidelines definition and in clinical management practice in order to increase hospital’s effectiveness and efficiency.