Measuring the Response Performance of U.S. States against COVID-19 Using an Integrated DEA, CART, and Logistic Regression Approach

Adapting Efficiency Analysis in Health Systems: A Scoping Review of Data Envelopment Analysis Applications During the COVID-19 Pandemic

Objective: To synthesize the current evidence base concerning the application of Data Envelopment Analysis (DEA) in healthcare efficiency during the COVID-19 pandemic using a scoping review of 13 primary studies. Methods: We consulted databases including Web of Science (WoS) and Scopus, as well as manual search entries up to September 2022. Included studies were primary applications of DEA for assessing healthcare efficiency during the COVID-19 pandemic. Key findings derived from thematic analysis of repeating pattern observations were extracted and tabulated for further synthesis, taking into consideration the variations in DMU definitions, the inclusion of undesirable outputs, the influence of external factors, and the infusion of advanced technologies in DEA. Results: The review observed a diverse application of DMUs, ranging from healthcare supply chains to entire national health systems. There was an evident shift towards incorporating undesirable outputs, such as mortality rates, in the DEA models amidst the pandemic. The influence of external and non-discretionary factors became more pronounced in DEA applications, highlighting the interconnected nature of global health challenges. Notably, several studies integrated advanced computational methods, including machine learning, into traditional DEA, paving the way for enhanced analytical capabilities. Conclusions: DEA, as an efficiency analysis tool, has exhibited adaptability and evolution in its application in the context of the COVID-19 healthcare crisis. By recognizing the multifaceted challenges posed by the pandemic, DEA applications have grown more comprehensive, integrating broader societal and health outcomes. This review provides pivotal insights that can inform policy and healthcare strategies, underscoring the importance of dynamic and comprehensive efficiency analysis methodologies during global health emergencies.

The efficiency evaluation of traditional Chinese medicine hospitals by data envelopment analysis in Zhengzhou, China

Aim

This study aimed to evaluate the operational efficiency of traditional Chinese medicine (TCM) hospitals in China.

Methods

Pearson's analysis was used to test the correlation between the input and output variables. Data envelopment analysis (DEA) was utilized to analyze the input and output variables of 16 TCM hospitals, and each hospital efficiency score was computed by Deap 2.1, assuming variable return to scale (VRS), which is an input-oriented model. t tests were conducted to confirm the significant difference of efficiency scores at the hospital level and by hospital type, and ANOVA was used to test for significant differences in efficiency scores according to hospitals' size.

Results

The correlation coefficient of the input and output indicators was between 0.613 and 0.956 (p < 0.05). The difference in number of doctors (ND) and numbers of pharmacists (NP) were statistically significant (p < 0.05) at the hospital level. The mean efficiency scores for technical efficiency (TE), pure technical efficiency (PTE), and scale efficiency (SE) in secondary TCM hospitals were 0.766, 0.919, and 0.838, respectively. Additionally, the lowest TE, PTE, and SE were 0.380, 0.426, and 0.380, respectively. Eight TCM hospitals in this study were DEA efficient, with an efficiency score of 1. There were no statistically significant differences in TE, PTE, and SE among hospital levels, hospital types or hospital sizes groups (p > 0.05).

Conclusion

This study revealed that tertiary TCM hospitals had a greater level of efficiency than secondary TCM hospitals. In our study, 50% of TCM hospitals had inefficient management. Therefore, to activate the new development power of TCM hospitals, it is necessary to reform and improve the management system and mechanism of TCM hospitals, optimize the development environment of TCM hospitals and formulate development plans and measures based on local conditions.

Predicting Nurse Turnover for Highly Imbalanced Data Using the Synthetic Minority Over-Sampling Technique and Machine Learning Algorithms

Predicting nurse turnover is a growing challenge within the healthcare sector, profoundly impacting healthcare quality and the nursing profession. This study employs the Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance issues in the 2018 National Sample Survey of Registered Nurses dataset and predict nurse turnover using machine learning algorithms. Four machine learning algorithms, namely logistic regression, random forests, decision tree, and extreme gradient boosting, were applied to the SMOTE-enhanced dataset. The data were split into 80% training and 20% validation sets. Eighteen carefully selected variables from the database served as predictive features, and the machine learning model identified age, working hours, electric health record/electronic medical record, individual income, and job type as important features concerning nurse turnover. The study includes a performance comparison based on accuracy, precision, recall (sensitivity), F1-score, and AUC. In summary, the results demonstrate that SMOTE-enhanced random forests exhibit the most robust predictive power in the classical approach (with all 18 predictive variables) and an optimized approach (utilizing eight key predictive variables). Extreme gradient boosting, decision tree, and logistic regression follow in performance. Notably, age emerges as the most influential factor in nurse turnover, with working hours, electric health record/electronic medical record usability, individual income, and region also playing significant roles. This research offers valuable insights for healthcare researchers and stakeholders, aiding in selecting suitable machine learning algorithms for nurse turnover prediction.

Measuring efficiency of public hospitals under the impact of Covid-19: the case of Türkiye

The Covid-19 pandemic has had serious medical, administrative and financial effects on the health system and hospitals around the world. In Türkiye, compared to 2019 realizations, in 2020 and 2021 respectively there were 39% and 21% decrease in the number of outpatient services and 29% and 17% decline in total inpatient services of public hospitals. The main subject of this research is how the pandemic period affects the Turkish public hospitals' efficiency. We have measured the technical efficiency of outpatient and inpatient care services of Turkish public hospitals using Stochastic Frontier Analysis (SFA). The dataset includes 563 hospitals for the years 2015 through 2021. Inputs of number of physicians, nurses and other medical staff, and number of beds and their interactions with each other are introduced to the SFA models of outputs of outpatient visits and inpatient discharges adjusted with case mix index are derived. Firstly, we found that the years associated with Covid-19 have a significant negative impact on the inpatient service efficiency. Training and Research and City Hospitals have low efficiency scores in outpatient services but high efficiency scores in inpatient services. In addition, the regions with high population rates have positive impact in outpatient efficiency and negative impact in inpatient efficiency. During the pandemic, city hospitals, have received large investments, gained a key role by increasing both the patient load and their efficiency. Future reforms can be guided by taking advantage of the efficiency differences of hospitals in different environmental factors.

Uncertain Malmquist productivity index: An application to evaluate healthcare systems during COVID-19 pandemic

Evaluation of healthcare systems, as a key organization providing different health services, is essential. This issue becomes more crucial when occurring crises such as a pandemic. They need to keep track of their success in the face of the crisis to assess the effects of policy changes and their capability to respond to new challenges. The Malmquist Productivity Index (MPI) is measured to analyze the causes of productivity change between two periods of time. The estimation of the traditional MPI requires reliable and detailed information on the inputs and outputs of decision-making units. However, there are a lot of situations where input and/or output may be imprecise. It is not manageable to reliably measure certain measurement indices, such as quality of treatment or system flexibility. For such cases, experts are invited to model their opinion. Uncertainty theory is a mathematical branch rationally dealing with belief degrees. The primary objective of this study is to apply MPI concept in the nonparametric approach of data envelopment analysis to calculate the efficiency of systems over different periods of time under uncertain conditions. Accordingly, we consider the MPI when inputs and outputs are belief degrees of experts. Furthermore, the sensitivity of the model is analyzed to determine the reliability of the results to the variation of variables. Finally, as an illustrative example, we explore longitudinal efficiency of healthcare systems during COVID-19 pandemic. According to the results of our model, the majority of the countries have improved in the second period which can be the result of efforts to improve pandemic preparedness. The decomposition of MPI into efficiency changes and technical changes indicates that the rise in productivity is entirely related to the progressive change of the production frontier related to policymaking. This application attempts to demonstrate how crucial it is to take uncertainties into account when comparing the performance of different systems over periods of time. The developed model enables us to consider the uncertainty existing in COVID-19 pandemic. The proposed model can handle more accurately the uncertainty during the pandemic. Thus, the result could be more reliable, which can benefit decision-makers in regard to performance improvement.

Efficiency Measurement Using Data Envelopment Analysis (DEA) in Public Healthcare: Research Trends from 2017 to 2022

With the shifting healthcare environment, the importance of public healthcare systems is being emphasized, and the efficiency of public healthcare systems has become a critical research agenda. We reviewed recent research on the efficiency of public healthcare systems using DEA, which is one of the leading methods for efficiency analysis. Through a systematic review, we investigated research trends in terms of research purposes, specific DEA techniques, input/output factors used for models, etc. Based on the review results, future research directions are suggested. The results of this paper provide valuable information and guidelines for future DEA research on public healthcare systems.

Efficiency of Malaysian states in managing the COVID-19 outbreak in 2020 and 2021

INTRODUCTION

Many developing countries have drastically imbalanced health systems in different regions. The COVID-19 outbreak posed a further challenge as hospital structures, equipped with doctors, critical care units and respirators, were not available to a sufficient extent in all regions.

OBJECTIVE

This study is a descriptive study on the efficiency of Malaysian states in facing the COVID-19 outbreak.

METHODOLOGY

The efficiency of all Malaysian states was measured using Data Envelopment Analysis in which each state's Score of COVID Index (SCI) was quantified. The SCI of these states were then further compared between the year 2020 and 2021. A greater disparity would indicate a decline in the performance of a state over time, where nearly all the states in Malaysia experienced an increase in the score of COVID Index (SCI).

RESULT

This study found that the central region was the most affected, since all the three states in the region (Selangor, Federal Territory of Kuala Lumpur, and Federal Territory of Putrajaya) showed a situation of inadequacy (SCI: >0.75) due to the COVID-19 outbreak.

CONCLUSION

The ranking of Malaysia's states according to their vulnerability to an outbreak of COVID-19 is vitally significant for the purposes of assisting the government and policymakers in planning their responses to the outbreak and ensuring that resources are distributed appropriately.

Emergency medical resource allocation among hospitals with non-regressive production technology: A DEA-based approach

This paper proposes an approach for medical resource allocation among hospitals under public health emergencies based on data envelopment analysis (DEA). First, the DEA non-regressive production technology is adopted to ensure that the DMU can always refer to the most advanced production technology throughout all production periods. Based on the non-regressive production technology, two efficiency evaluation models are presented to calculate the efficiencies of DMUs before and after resource allocation. Our theoretical analysis shows that all the DMUs can be efficient after medical resource allocation, and thus a novel resource allocation possibility set is developed. Further, two objectives are considered and a bi-objective resource allocation model is developed. One objective is to maximize the output target realizability of the DMUs, while the other is to ensure the allocated resource to each DMU fits with its operation size, preperformance, and operation practice (i.e., proportion of critically ill patients). Additionally, a trade-off model is proposed to solve the bi-objective model to obtain the final resource allocation results. The proposed approach contributes by ensuring that the medical resources are allocated in such a way that they can all be efficiently used as well as considering multiple objectives and practical constraints that make the approach more fitted with the practical application scenarios. Finally, a case study of 30 hospitals in Wuhan during the COVID-19 epidemic is applied to illustrate the proposed approach.

The Core of Healthcare Efficiency: A Comprehensive Bibliometric Review on Frontier Analysis of Hospitals

Parametric and non-parametric frontier applications are typical for measuring the efficiency and productivity of many healthcare units. Due to the current COVID-19 pandemic, hospital efficiency is the center of academic discussions and the most desired target for many public authorities under limited resources. Investigating the state of the art of such applications and methodologies in the healthcare sector, besides uncovering strategical managerial prospects, can expand the scientific knowledge on the fundamental differences among efficiency models, variables and applications, drag research attention to the most attractive and recurrent concepts, and broaden a discussion on the specific theoretical and empirical gaps still to be addressed in future research agendas. This work offers a systematic bibliometric review to explore this complex panorama. Hospital efficiency applications from 1996 to 2022 were investigated from the Web of Science base. We selected 65 from the 203 most prominent works based on the Core Publication methodology. We provide core and general classifications according to the clinical outcome, bibliographic coupling of concepts and keywords highlighting the most relevant perspectives and literature gaps, and a comprehensive discussion of the most attractive literature and insights for building a research agenda in the field.

Application of machine learning to a material library for modeling of relationships between material properties and tablet properties

This study investigates the usefulness of machine learning for modeling complex relationships in a material library. We tested 81 types of active pharmaceutical ingredients (APIs) and their tablets to construct the library, which included the following variables: 20 types of API material properties, one type of process parameter (three levels of compression pressure), and two types of tablet properties (tensile strength (TS) and disintegration time (DT)). The machine learning algorithms boosted tree (BT) and random forest (RF) were applied to analysis of our material library to model the relationships between input variables (material properties and compression pressure) and output variables (TS and DT). The calculated BT and RF models achieved higher performance statistics compared with a conventional modeling method (i.e., partial least squares regression), and revealed the material properties that strongly influence TS and DT. For TS, true density, the tenth percentile of the cumulative percentage size distribution, loss on drying, and compression pressure were of high relative importance. For DT, total surface energy, water absorption rate, polar surface energy, and hygroscopicity had significant effects. Thus, we demonstrate that BT and RF can be used to model complex relationships and clarify important material properties in a material library.