Predictors of hospitalization length of stay among re-admitted treatment-resistant Bipolar Disorder inpatients

Fornaro, Michele; Iasevoli, Felice; Novello, Stefano; Fusco, Andrea; Anastasia, Annalisa; Berardis, Domenico De; Valchera, Alessandro; Bartolomeis, Andrea de

doi:10.1016/j.jad.2017.12.009

Cited by 11 publications

(7 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There have been studied the factors affecting the LOS in psychiatric patients. [ 21 , 22 ] As psychiatric patients have different hospitalization factors according to personal characteristics, it seems to be needed to find effective ways of LOS management by analyzing various personal characteristics and treatment pathways among psychiatric patients.…”

Section: Discussionmentioning

confidence: 99%

Analysis of length of hospital stay using electronic health records: A statistical and data mining approach

et al. 2018

View full text Add to dashboard Cite

BackgroundThe length of stay (LOS) is an important indicator of the efficiency of hospital management. Reduction in the number of inpatient days results in decreased risk of infection and medication side effects, improvement in the quality of treatment, and increased hospital profit with more efficient bed management. The purpose of this study was to determine which factors are associated with length of hospital stay, based on electronic health records, in order to manage hospital stay more efficiently.Materials and methodsResearch subjects were retrieved from a database of patients admitted to a tertiary general university hospital in South Korea between January and December 2013. Patients were analyzed according to the following three categories: descriptive and exploratory analysis, process pattern analysis using process mining techniques, and statistical analysis and prediction of LOS.ResultsOverall, 55% (25,228) of inpatients were discharged within 4 days. The department of rehabilitation medicine (RH) had the highest average LOS at 15.9 days. Of all the conditions diagnosed over 250 times, diagnoses of I63.8 (cerebral infarction, middle cerebral artery), I63.9 (infarction of middle cerebral artery territory) and I21.9 (myocardial infarction) were associated with the longest average hospital stay and high standard deviation. Patients with these conditions were also more likely to be transferred to the RH department for rehabilitation. A range of variables, such as transfer, discharge delay time, operation frequency, frequency of diagnosis, severity, bed grade, and insurance type was significantly correlated with the LOS.ConclusionsAccurate understanding of the factors associating with the LOS and progressive improvements in processing and monitoring may allow more efficient management of the LOS of inpatients.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of length of hospital stay using electronic health records: A statistical and data mining approach

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Previous studies (e.g. [ 6 – 9 ]) have done important work in identifying predictors (e.g. quality of life, psychiatric diagnosis, impact of symptoms, living situation) that can be relevant for a prognostic model for psychiatric hospitalization.…”

Section: Introductionmentioning

confidence: 99%

Predicting hospitalization following psychiatric crisis care using machine learning

Blankers

Post

Dekker

2020

BMC Med Inform Decis Mak

View full text Add to dashboard Cite

Background Accurate prediction models for whether patients on the verge of a psychiatric criseis need hospitalization are lacking and machine learning methods may help improve the accuracy of psychiatric hospitalization prediction models. In this paper we evaluate the accuracy of ten machine learning algorithms, including the generalized linear model (GLM/logistic regression) to predict psychiatric hospitalization in the first 12 months after a psychiatric crisis care contact. We also evaluate an ensemble model to optimize the accuracy and we explore individual predictors of hospitalization. Methods Data from 2084 patients included in the longitudinal Amsterdam Study of Acute Psychiatry with at least one reported psychiatric crisis care contact were included. Target variable for the prediction models was whether the patient was hospitalized in the 12 months following inclusion. The predictive power of 39 variables related to patients’ socio-demographics, clinical characteristics and previous mental health care contacts was evaluated. The accuracy and area under the receiver operating characteristic curve (AUC) of the machine learning algorithms were compared and we also estimated the relative importance of each predictor variable. The best and least performing algorithms were compared with GLM/logistic regression using net reclassification improvement analysis and the five best performing algorithms were combined in an ensemble model using stacking. Results All models performed above chance level. We found Gradient Boosting to be the best performing algorithm (AUC = 0.774) and K-Nearest Neighbors to be the least performing (AUC = 0.702). The performance of GLM/logistic regression (AUC = 0.76) was slightly above average among the tested algorithms. In a Net Reclassification Improvement analysis Gradient Boosting outperformed GLM/logistic regression by 2.9% and K-Nearest Neighbors by 11.3%. GLM/logistic regression outperformed K-Nearest Neighbors by 8.7%. Nine of the top-10 most important predictor variables were related to previous mental health care use. Conclusions Gradient Boosting led to the highest predictive accuracy and AUC while GLM/logistic regression performed average among the tested algorithms. Although statistically significant, the magnitude of the differences between the machine learning algorithms was in most cases modest. The results show that a predictive accuracy similar to the best performing model can be achieved when combining multiple algorithms in an ensemble model.

show abstract

“…Finally, the AUC of the model in this study ranged from 0.553 to 0.619, which is close to the random model. This may be due to insufficient variables covered in the study, such as quantitative measures of symptom severity, electroconvulsive therapy (Patel et al, 2019(Patel et al, ), 10.3389/fnhum.2023 psychotropic medication type and dosage (Deng et al, 2018;Fornaro et al, 2018;Lee et al, 2020), and health insurance (Masciale et al, 2021).…”

Section: Limitationmentioning

confidence: 99%

Demographic, clinical and biochemical correlates of the length of stay for different polarities in Chinese inpatients with bipolar disorder: A real-world study

Wang

Du²,

Li³

et al. 2023

Front. Hum. Neurosci.

View full text Add to dashboard Cite

Introduction: Many patients with bipolar disorder (BD) need hospitalization, while the number of hospital beds for these patients is limited. Managing the length of stay (LOS) is an effective solution to this issue. Research on LOS and its influencing factors in BD is limited in China. This study aimed to identify the factors relevant to LOS in different polarities in Chinese patients with BD.Method: This was a real-world, cross-sectional study. Data were obtained from the electronic medical record system. Patients admitted to Beijing Anding Hospital between Jan 2014 and Dec 2017 and diagnosed with BD were included. Demographic information, clinical characteristics, and biochemical variables were collected. Patients were classified into short and long LOS groups based on a cutoff value. A univariate study and a multivariate logistic regression analysis were performed to identify variables related to LOS in various BD polarities. The receiver operating characteristic (ROC) analysis was utilized to evaluate the discrimination accuracy of the regression model.Result: Four thousand six hundred and seventy-five visits from 4,451 individuals were included in the analysis. For the whole sample, unmarried status, psychotic features, and family history of mental disorders were positively associated with long LOS (all p < 0.05). There was an additive interaction between a family history of mental disorders and polarities (p < 0.05). For manic episodes, unmarried status, psychotic features, and family history of mental disorders were positively associated with long LOS (all p < 0.05). For depressive episodes, psychotic features and high-density lipoprotein cholesterol (HDLC) levels were positively associated with long LOS (all p < 0.05). For mixed states, unmarried status was positively associated with long LOS, while low-density lipoprotein cholesterol (LDLC) levels were negatively associated with LOS (all p < 0.05). The area under the curve (AUC) values for depressive episodes, manic episodes, and mixed states in the combined model were 0.587, 0.553, and 0.619, respectively (all p < 0.05).Discussion: The findings suggested that LOS correlates differed by polarity, with marital status, psychotic features, a family history of mental disorders, and lipid levels strongly linked with LOS in patients with BD.

show abstract

Predictors of hospitalization length of stay among re-admitted treatment-resistant Bipolar Disorder inpatients

Cited by 11 publications

References 42 publications

Analysis of length of hospital stay using electronic health records: A statistical and data mining approach

Analysis of length of hospital stay using electronic health records: A statistical and data mining approach

Predicting hospitalization following psychiatric crisis care using machine learning

Demographic, clinical and biochemical correlates of the length of stay for different polarities in Chinese inpatients with bipolar disorder: A real-world study

Contact Info

Product

Resources

About