Extract critical factors affecting the length of hospital stay of pneumonia patient by data mining (case study: an Iranian hospital)

Unveiling the Negative Customer Experience in Diagnostic Centers: A Data Mining Approach

Introduction

This study aims to identify the negative customer experiences reflected in complaints against diagnostic centers using data mining tools.

Methods

Analyzing customer complaints from a consumer complaints website, the Apriori algorithm was employed to uncover frequent patterns and identify key areas of concern. The frequency and distribution of terms used in complaints were also analyzed, and word clouds were generated to visualize the findings.

Results

The study revealed that major areas of unfavorable customer experience included delayed test reports, erroneous test results, difficulties scheduling appointments, staff incivility, subpar service, and medical negligence.

Discussion

These findings and the proposed model can guide diagnostic centers in incorporating data mining tools for customer experience analysis, enabling managers to proactively address issues and view complaints as opportunities for service improvement rather than legal liabilities.

A Comprehensive Overview of IoT-Based Federated Learning: Focusing on Client Selection Methods

The integration of the Internet of Things (IoT) with machine learning (ML) is revolutionizing how services and applications impact our daily lives. In traditional ML methods, data are collected and processed centrally. However, modern IoT networks face challenges in implementing this approach due to their vast amount of data and privacy concerns. To overcome these issues, federated learning (FL) has emerged as a solution. FL allows ML methods to achieve collaborative training by transferring model parameters instead of client data. One of the significant challenges of federated learning is that IoT devices as clients usually have different computation and communication capacities in a dynamic environment. At the same time, their network availability is unstable, and their data quality varies. To achieve high-quality federated learning and handle these challenges, designing the proper client selection process and methods are essential, which involves selecting suitable clients from the candidates. This study presents a comprehensive systematic literature review (SLR) that focuses on the challenges of client selection (CS) in the context of federated learning (FL). The objective of this SLR is to facilitate future research and development of CS methods in FL. Additionally, a detailed and in-depth overview of the CS process is provided, encompassing its abstract implementation and essential characteristics. This comprehensive presentation enables the application of CS in diverse domains. Furthermore, various CS methods are thoroughly categorized and explained based on their key characteristics and their ability to address specific challenges. This categorization offers valuable insights into the current state of the literature while also providing a roadmap for prospective investigations in this area of research.

Early Determinants of Length of Hospital Stay: A Case Control Survival Analysis among COVID-19 Patients admitted in a Tertiary Healthcare Facility of East India

BACKGROUND

Length of hospital stay (LOS) for a disease is a vital estimate for healthcare logistics planning. The study aimed to illustrate the effect of factors elicited on arrival on LOS of the COVID-19 patients.

MATERIALS AND METHODS

It was a retrospective, record based, unmatched, case control study using hospital records of 334 COVID-19 patients admitted in an East Indian tertiary healthcare facility during May to October 2020. Discharge from the hospital (cases/survivors) was considered as an event while death (control/non-survivors) as right censoring in the case-control survival analysis using cox proportional hazard model.

RESULTS

Overall, we found the median LOS for the survivors to be 8 days [interquartile range (IQR): 7-10 days] while the same for the non-survivors was 6 days [IQR: 2-11 days]. In the multivariable cox-proportional hazard model; travel distance (>16 km) [adjusted hazard ratio (aHR): 0.69, 95% CI: (0.50-0.95)], mode of transport to the hospital (ambulance) [aHR: 0.62, 95% CI: (0.45-0.85)], breathlessness (yes) [aHR: 0.56, 95% CI: (0.40-0.77)], number of co-morbidities (1-2) [aHR: 0.66, 95% CI: (0.47-0.93)] (≥3) [aHR: 0.16, 95% CI: (0.04-0.65)], COPD/asthma (yes) [ [aHR: 0.11, 95% CI: (0.01-0.79)], DBP (<60/≥90) [aHR: 0.55, 95% CI: (0.35-0.86)] and qSOFA score (≥2) [aHR: 0.33, 95% CI: (0.12-0.92)] were the significant attributes affecting LOS of the COVID-19 patients.

CONCLUSION

Factors elicited on arrival were found to be significantly associated with LOS. A scoring system inculcating these factors may be developed to predict LOS of the COVID-19 patients.

Monitoring Diagnostic Safety Risks in Emergency Departments: Protocol for a Machine Learning Study

BACKGROUND

Diagnostic decision making, especially in emergency departments, is a highly complex cognitive process that involves uncertainty and susceptibility to errors. A combination of factors, including patient factors (eg, history, behaviors, complexity, and comorbidity), provider-care team factors (eg, cognitive load and information gathering and synthesis), and system factors (eg, health information technology, crowding, shift-based work, and interruptions) may contribute to diagnostic errors. Using electronic triggers to identify records of patients with certain patterns of care, such as escalation of care, has been useful to screen for diagnostic errors. Once errors are identified, sophisticated data analytics and machine learning techniques can be applied to existing electronic health record (EHR) data sets to shed light on potential risk factors influencing diagnostic decision making.

OBJECTIVE

This study aims to identify variables associated with diagnostic errors in emergency departments using large-scale EHR data and machine learning techniques.

METHODS

This study plans to use trigger algorithms within EHR data repositories to generate a large data set of records that are labeled trigger-positive or trigger-negative, depending on whether they meet certain criteria. Samples from both data sets will be validated using medical record reviews, upon which we expect to find a higher number of diagnostic safety events in the trigger-positive subset. Machine learning will be used to evaluate relationships between certain patient factors, provider-care team factors, and system-level risk factors and diagnostic safety signals in the statistically matched groups of trigger-positive and trigger-negative charts.

RESULTS

This federally funded study was approved by the institutional review board of 2 academic medical centers with affiliated community hospitals. Trigger queries are being developed at both organizations, and sample cohorts will be labeled using the triggers. Machine learning techniques such as association rule mining, chi-square automated interaction detection, and classification and regression trees will be used to discover important variables that could be incorporated within future clinical decision support systems to help identify and reduce risks that contribute to diagnostic errors.

CONCLUSIONS

The use of large EHR data sets and machine learning to investigate risk factors (related to the patient, provider-care team, and system-level) in the diagnostic process may help create future mechanisms for monitoring diagnostic safety.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/24642.

A study of factors related to patients' length of stay using data mining techniques in a general hospital in southern Iran

Purpose

The length of stay (LOS) in hospitals is a widely used indicator for goals such as health care management, quality control, utilizing hospital services and resources, and determining the degree of efficiency. Various methods have been used to identify the factors influencing the LOS. This study adopts a comparative approach of data mining techniques for investigating effective factors and predict the length of stay in Shahid-Mohammadi Hospital, Bandar Abbas, Iran.

Methods

Using a dataset consists of 526 patient records of the Shahid-Mohammadi Hospital from March 2016 to March 2017, factors affecting the LOS were ranked using information gain and correlation indices. In addition, classification models for LOS prediction were created based on nine data mining classifiers applied with and without feature selection technique. Finally, the models were compared.

Results

The most important factors affecting LOS are the number of para-clinical services, counseling frequency, clinical ward, the specialty and the degree of the doctor, and the cause of hospitalization. In addition, regarding to the classifiers created based on the dataset, the best accuracy (83.91%) and sensitivity (80.36%) belongs to the Logistic Regression and Naïve Bayes respectively. In addition, the best AUC (0.896) belongs to the Random Forest and Generalized Linear classifiers.

Conclusion

The results showed that most of the proposed models are suitable for classification of the length of stay, although the Logistic Regression might have a slightly better performance than others in term of accuracy, and this model can be used to determine the patients' Length of Stay. In general, continuous monitoring of the factors influencing each of the performance indicators based on proper and accurate models in hospitals is important for helping management decisions.

Predicting hospital-acquired pneumonia among schizophrenic patients: a machine learning approach

Background

Medications are frequently used for treating schizophrenia, however, anti-psychotic drug use is known to lead to cases of pneumonia. The purpose of our study is to build a model for predicting hospital-acquired pneumonia among schizophrenic patients by adopting machine learning techniques.

Methods

Data related to a total of 185 schizophrenic in-patients at a Taiwanese district mental hospital diagnosed with pneumonia between 2013 ~ 2018 were gathered. Eleven predictors, including gender, age, clozapine use, drug-drug interaction, dosage, duration of medication, coughing, change of leukocyte count, change of neutrophil count, change of blood sugar level, change of body weight, were used to predict the onset of pneumonia. Seven machine learning algorithms, including classification and regression tree, decision tree, k-nearest neighbors, naïve Bayes, random forest, support vector machine, and logistic regression were utilized to build predictive models used in this study. Accuracy, area under receiver operating characteristic curve, sensitivity, specificity, and kappa were used to measure overall model performance.

Results

Among the seven adopted machine learning algorithms, random forest and decision tree exhibited the optimal predictive accuracy versus the remaining algorithms. Further, six most important risk factors, including, dosage, clozapine use, duration of medication, change of neutrophil count, change of leukocyte count, and drug-drug interaction, were also identified.

Conclusions

Although schizophrenic patients remain susceptible to the threat of pneumonia whenever treated with anti-psychotic drugs, our predictive model may serve as a useful support tool for physicians treating such patients.