“The impact of DRG-based management of healthcare facilities on amenable mortality in the European Union”

ARTICLE INFO Rastislav Briestensky and Aleksandr Ključnikov (2021). The impact of DRGbased management of healthcare facilities on amenable mortality in the European Union. Problems and Perspectives in Management, 19(2), 264-275. doi:10.21511/ppm.19(2).2021.22 DOI http://dx.doi.org/10.21511/ppm.19(2).2021.22 RELEASED ON Thursday, 10 June 2021 RECEIVED ON Wednesday, 24 March 2021 ACCEPTED ON Thursday, 03 June 2021


INTRODUCTION
For a proper analysis of the efficiency of healthcare financing and thus the effectiveness of hospital management in the EU countries, it is essential to evaluate the general environment and its specific regional characteristics. The national healthcare systems generally differ in several ways. The primary classification comes from the OECD (Organization for Economic Co-operation and Development). According to the method of financing, the OECD divides health systems as follows: voluntary insurance, social health insurance, and universal coverage. In addition, they add two other types regarding the current situation: compulsory national health insurance and residual programs. It is necessary to postpone the classical classification and create a classification based on the segmentation of healthcare services or the population. Based on this idea, Toth (2016) designed ten different models of healthcare systems. Each healthcare system has its different specifics, and therefore the selected groups of states should be taken into consideration when analyzing the current state or proposing any solution.
Hospitals are open systems that operate dynamically outside the equilibrium point. Therefore, it is often possible to observe unpredictable environmental behavior (Diaz & Castilla, 2017). Hospitals can be divided based on several criteria: by ownership -public and private hospitals; by financing -focus on profitability or non-profit; by educational activities -faculty and non-pedagogical hospitals; by hierarchical classification -primary, secondary and tertiary hospitals; by the degree of specialization -general and specialized hospitals; by the status of employees and physicians -hospital and personnel model (Asbu et al., 2020). There are many differences between hospitals, and each of them is somehow specific. Managers must understand these specifics to define the most suitable financing strategy. They may apply general procedures but should consider the specific characteristics of each hospital in the strategy.
When analyzing standard costing methods for individual hospital wards in Europe, the lack of standardized methodologies for determining the exact costs of hospital wards can be pointed out. A standardized costing methodology would facilitate comparisons, streamline economic evaluation within the department, and assist in the decision-making process regarding efficient resource allocation (Negrini et al., 2004). Each country in the European Union has a different system of healthcare financing. Several financing models are described in the literature: the Bismarck model, the Beveridge model, the National healthcare delivery model (a combination of the Bismarck model and the Beveridge model), and the Out-of-pocket model (Khaled & Nessef, 2018). Since the 1990s, DRG-based (diagnosis-related group) management and financing was introduced in many countries to improve the efficiency of financial resources (Mihailovic et al., 2016).
Healthcare reforms force hospital managers to focus on the efficiency of financial resources. By evaluating resource efficiency, they can determine their performance (Giménez et al., 2019). It is estimated that there is a connection between hospital performance and the quality of health care. Efficiency is based on produced outputs concerning inputs (Schreyögg, 2019).
After evaluating performance, it is possible to reveal the strengths and weaknesses of each hospital or selected healthcare system (Setiawannie & Rahmania, 2019). By finding this information, it is possible to compare individual hospitals and countries based on the selected criteria. In this way, it is possible to determine which of the factors influence hospital performance the most (Bhaduri, 2020). Amenable mortality, defined as premature death that can be avoided through the optimal quality of healthcare, is one of the essential outputs of the healthcare system (Gianino et al., 2017). There is a relationship between amenable mortality and the quality of healthcare. Amenable mortality is one of the indicators of healthcare quality (Kruk et al., 2018). The European Commission monitors it among the Member States to evaluate the healthcare system performance of a country (Weber & Clerc, 2017).
In most cases, the influence of DRG is evaluated based on the selected factors of the general environment, transparency, and financial efficiency. Nevertheless, there is not enough evidence about the direct influence of DRG on medical performance indicators, such as amenable mortality in the European Union countries.

LITERATURE REVIEW
The hospital financing system is considered an essential part of their management (Gaspar et al., 2020). Among other shortcomings, some countries in the EU do not apply a comprehensive approach to their financial management. Given the above, sufficient reforms are needed in these areas. One of the possible changes can be the introduc-tion of DRG-based management and financing (Dubas-Jakóbczyk et. al, 2020). DRG-based financing classifies each patient's case according to the diagnosis and other characteristics. Payment for treatment is therefore based on DRG groups and used resources (WHO, 2012).
DRGs are considered to be tools for hierarchical cost control within the US private health in-surance system. At the same time, their application in the English national health service system may increase the productivity of hospital services. Additionally, it is established that DRGs boost competition within the German social health insurance system, and thus are viewed as tools for managing self-regulated providers. In this regard, DRG stimulates hybridization of a healthcare sector contributing to the update of less developed mechanisms that already exist in such a sector (Schmid et al., 2010). DRG-based financing systems provide better measurability of healthcare costs (European Commission, 2017). Better measurability brings more control over expenditures, as well as their possible direction in the desired way. Subsequently, managers can identify the strengths and weaknesses of financing, leading to a set of recommendations for streamlining financing (Tan et al., 2012). For this reason, it is necessary to distinguish healthcare between countries that have a DRG-based financing system and those that do not use such a system.
As each country usually uses a unique DRG system, it is crucial to know the purpose for which the whole system is designed. The professional domestic and foreign literature contains an analysis of each of these countries, from financing strategies to operational activities (Quentin et al., 2013). It is sufficient to understand which countries have established DRG financing systems, from which year, and for what purposes to assess the effectiveness of countries. DRG-related financing is used in the If any country wants to introduce DRG-based management and financing, it can choose from the existing systems or create a specific version of the system (Kashilska & Petkov, 2019). Each of the existing systems has its criteria of classification (Talaghir et al., 2018). As each country has different ways of providing health care and the structure of providers, it must choose specific criteria when implementing DRG (Milcent, 2021). Each country should make a separate study to set case weights to implement it; otherwise, it can create the wrong incentives (Mathauer & Wittenbecher, 2013). According to these findings, it is possible to conclude that the implementation of this system is a process that requires proper setup. Therefore, when implementing this method of financing, it is essential to examine the state of healthcare providing of each country carefully, and the results can be hardly generalized (Stephani et al., 2017) Every financing system implemented across the countries of the European Union has its strengths and weaknesses. DRG-based management and financing are applied across many countries, as it has many advantages such as financial transparency or the possibility to evaluate hospital performance. The application of the DRG system presupposes increased efficiency, and a reduction in the length of hospital stay (Mihailovic et al., 2016), transparent and fair allocation of financial resources, equal payments for equally demanding hospitalizations, higher payment for more demanding hospitalizations, unification of contractual relations between health insurance companies and health care providers, higher motivation to keep documentation on provided healthcare services and the possibility of measuring and comparing the productivity of hospitals (Ozorovský, 2016). DRG sets a cost ceiling beneath so that healthcare providers can reduce their costs (Sheaff et al., 2020). Nevertheless, this system can provide an incentive to transfer patients to home treatment sooner than if it was not in place. Therefore, the application of DRG may not have a clear medical benefit (Mihailovic et al., 2016). It can also be stated that DRG application does not always bring an increase in system efficiency (Cylus et al., 2017). When managers want to evaluate the efficiency of this system, they should consider the relationship between efficiency, equality, and equity (Chletsos & Saiti, 2019).
Each country can have different characteristics that must be considered when comparing them. Medical tourism is one of the important ones. In terms of medical tourism, patients travel to other countries for medical procedures. Patients may be motivated by one or more of the following factors: lower prices, higher quality of services, the performance of specific actions not available in their own country, the possibility of connecting with holidays and convalescence in each country, etc. (Stewart, 2018). Health tourism can significantly impact measuring the performance of target countries, as these countries can't be included in the comparison. Data on the number of doctors, nurses, and beds per 1,000 inhabitants may be distorted within the statistics. Furthermore, this may affect financing, as these services are paid by patients and not by the hospital budget. Including these countries in the performance, comparison amends the results. According to several expert sources, Romania and Bulgaria can be considered among the countries most affected by health tourism ( Without measurement, it cannot be expected to improve hospital management quality to reach a high level (Quentin et al., 2019). In assessing effectiveness, the World Health Organisation (WHO) recommends focusing on inputs and outputs in healthcare. Therefore, it is a question of assessing whether, given the inputs to be interpreted, a particular country, region, or health facility achieves the optimal number of outputs. Based on the input-output-oriented analysis, the performance of health care systems can be measured. WHO methodology uses input-output evaluation of the effectiveness of hospitals (WHO, 2012). For this purpose, inputs and outputs must be selected to meet the objective of the research. Based on the above, it is possible to calculate hospital efficiency as a proportion of outputs to inputs. Different input and output data were used. The data was selected mainly regarding the group of countries studied or based on theoretical knowledge to be verified. To compare states or determine the effectiveness in finding the ideal model of functioning, only those factors that are applicable in the environment of the European Union countries were used. The selection of factors can be based on the WHO recommendation. It uses some standard indicators, usually recalculated per 1,000 inhabitants or assigning them to a specific age group of the population. All indicators are in a standardized form and thus consider the structure of the population based on age and gender, the structure of the population in other countries, or individual national statistics standardized differently. These indicators include in particular: age-standardized mortality rate per 100,000 population, life expectancy at birth and the age of 65, infant mortality, health expenditure per capita measures in PPP (purchasing-power-parity), number of doctors and nurses per 1,000 inhabitants, number of beds per 1,000 inhabitants, healthy years of life at the age of 65 (WHO, 2019).
Many authors presented the studies on this topic performed using Data Envelopment Analysis (DEA). Several studies work with different inputs and outputs, for example: number of beds, number of doctors, percentage of GDP spent on healthcare as inputs and life expectancy at birth, life expectancy adjusted for health, infant mortality rate as outputs (Asandului et al., 2014); number of beds, number of doctors as inputs and operating income, number of cases, days of hospitalization as outputs (Nistor et. al., 2017); number of beds, number of doctors, number of nurses, number of therapists as inputs and number of patients discharged during a given period as outputs (Dénes et. al

AIMS, HYPOTHESIS AND METHODOLOGY
Healthcare financing is a critical factor that influences hospital management. This paper focuses on the effectiveness of healthcare financing in countries with more funds but cannot manage them efficiently.
The aim is to determine whether there is a significant dependency between DRG-based (diagnosis-related groups) management of healthcare facilities and amenable mortality in the European Union countries. The literature review allows as-suming that DRG-based managed healthcare systems can be more efficient than others. Based on the literature review and the aim of the paper, the following hypothesis was formulated: H0: There is no statistically significant difference between the mean value of the order of countries with DRG-based managed healthcare systems and countries with the other unique healthcare financing system according to the selected model.
The alternative hypothesis states the following: H1: There is a statistically significant difference between the mean value of the order of countries with DRG-based managed healthcare systems and countries with other unique healthcare financing systems according to the selected model.
In this research, the hospital management approaches are to be examined at the level of individual countries, not at the level of individual facilities, because of the assumption that individual hospitals within the selected country will not have significant differences in the areas of financing and legislation.
The countries were divided into two groups for this research: 1. States with DRG-based healthcare management and financing and 2. States with unique healthcare financing systems. To improve the accuracy of the results, Bulgaria and Romania were excluded from the research sample, while the main reason was the extensive development of medical tourism.
Input data represent input values that affect the efficiency of hospital management. With a low number of beds, there is a higher waiting time for procedures and other negative consequences.
With an excessive number, the occupancy of hospitals is lower, which increases costs and reduces the efficiency of use. Based on the WHO methodology and the literature review, a unique combination of inputs and outputs was selected. Financial factors, human resources, and equipment were defined as the inputs, and such healthcare quality factors as mortality, life expectancy, and treatment time were considered as the outputs.
Focusing on the differences in the financial background of the selected states, health expenditure per capita measured in PPP was selected as the first input. The impact of the number of doctors and nurses involved in the healthcare system can also be significant. By examining the effectiveness, it is possible to determine which countries have the most effectively used medical staff. The number of physicians per 100,000 inhabitants and the number of nurses per 100,000 inhabitants were selected as the other inputs. The number of hospital beds per 1,000 inhabitants was selected as the last input to consider the differences in hospital resources, as it can differ in selected countries. Amenable mortality as a quality indicator was selected as the measured output.  (WHO, 2000). In this way, it helps to identify the factors associated with successful hospital management. The method can be input-oriented or output-oriented (Ahmed et al., 2019). In the case of hospitals, it helps to identify how the specific inputs affect the outputs. When comparing within the EU, it is possible to determine what change in inputs can be used to improve outputs and which countries work most effectively concerning the selected inputs and outputs. From the data obtained, it is possible to propose a set of basic recommendations for other countries, which will result in increased efficiency of hospital man-agement. The input-oriented version is used in the research to determine how to minimize inputs to achieve the desired outputs.
DEA analysis is based on the decision-making units (DMU). In this case, each DMU represents a particular country. Efficiency is calculated based on the data provided and is limited by it. The calculation of the DEA analysis is based on two steps: 1. The limit is determined; 2. The efficiency score obtained after comparing the inputs and outputs that are located at the most efficient DMU is assigned to each DMU. The ratio of outputs to inputs is used to calculate the efficiency of each DMU at a distance from the limit of the most efficient DMU. Assume that there are n DMUs, m inputs, and s outputs. The efficiency score of the tested DMU is obtained by calculating the proposed model. The above-presented approach is run n times to identify the relative efficacy scores of all DMUs. Each DMU selects input and output scales that maximize its efficiency score.
The Mann-Whitney U-test, a nonparametric equivalent of the t-test (MacFarland & Yates, 2016), is used with nonparametric data. Using this test, it can be determined whether there is a relationship between two dependent groups of data (Stehlíková, 2009).
There are considered measurements from data group A from the set 1

RESULTS AND DISCUSSION
Using DEA analysis and selected model of inputs (Health expenditure per capita measures in PPP, Number of physicians per 100,000 inhabitants, Number of nurses per 100,000 inhabitants and beds per 1,000 inhabitants) and output (Amenable mortality), thetaOpt values are calculated, which determine the degree of efficiency between selected countries using predefined input and output data. The higher the value, the better the efficiency. A higher thetaOpt value means that the country uses its resources more optimally. The maximum value of the thetaOpt is 1.
The country that reaches thetaOpt value 1 each year has the most optimal use of resources compared to other selected countries. DRG-based management and financing are focused on optimizing financing, and therefore the optimization of the input-output ratio is used instead of focusing only on output indicators. An overview of individual countries and the development of the situation for individual years together with data on the implementation or failure to implement DRG-based management and financing system are presented in Table 1. The value of thetaOpt was rounded to 3 decimal places.
After calculating thetaOpt value, a ranking of countries according to the optimality of their healthcare based on selected inputs and outputs for each year separately was created. The country with the highest value of thetaOpt is at the first place. The evaluation was performed for each year separately. Subsequently, countries were divided into two groups according to whether they have DRG-based management and financing systems in place. Two groups of countries with an assigned sequence number 1-26 were created. Figure 1 presents the value of thetaOpt in selected EU countries according to the selected model. It must be mentioned that there is a dependency between the performance (thetaOpt) and amenable mortality. Evaluation of performance shows which countries have better amenable mortality considering their inputs. Despite this, country values by amenable mortality can be different, e.g., Latvia, with the best performance of inputs and outputs, has the worst value of amenable mortality.
After performing DEA analysis, an order of all countries based on their performance with num-bers 1-26 for each year (the period of evaluation is 2012-2017) was made. The calculation for each year was performed separately. For each year, countries were divided into two groups by hospital financing: 1. States with DRG-based management and financing and 2. States with a unique financ-   ing system. After that, the Mann-Whitney U-Test was used to determine the existence of a significant difference in mean efficacy between the two groups. After its evaluation for individual years, Table 2 was created.
U represents the value of the test statistic, and p-value expresses statistical significance. Since p is greater than 0.05, the hypothesis H0 is confirmed, the alternative hypothesis H1 is rejected. The results give a significant finding -DRGbased healthcare management and financing system does not affect amenable mortality in the European Union countries. As already mentioned, several studies have shown the positive impact of the DRG on transparency, possibilities for evaluating financing, transparency in the use of funds spent, and other factors. However, this system did not bring any significant changes in the medical field in the form of the impact on amenable mortality. However, this does not mean that the application of this system is unnecessary or ineffective.
As each of the countries has a different healthcare system, it is not possible to analyze all of them in detail. Therefore, some factors can significantly impact strategy creation, and managers should focus on them. DEA analysis showed that there was no statistically significant difference between the effectiveness of countries with their financing systems and DRG-based management and financing focusing on deaths from treatable diseases. It can be assumed that DRG will have benefits in each established country. After increasing transparency, it will be easier for managers to manage financial flows, analyze them and suggest possible solutions.  When comparing thetaOpt values, which represent efficiency and amenable mortality, it is possible to point to another new finding. The country with the highest efficiency rate may not have the lowest amenable mortality. As countries increase the number of financial resources and other resources spent on healthcare, the rate of optimization decreases from a certain value. Based on the above, it is possible to explain, for example, the highest efficiency rate of Latvia, which also has the highest amenable mortality. Since Mihailovic et al.

CONCLUSION
The influence of DRG on the selected factors of the general environment, transparency, and financial efficiency was evaluated. Nevertheless, there is not enough evidence about the direct influence of DRG on medical performance indicators, such as amenable mortality in the European Union countries.
This article aimed to determine whether there is a significant dependency between DRG-based (diagnosis-related groups) management of healthcare facilities and amenable mortality in the European Union countries. Health expenditure per capita measured in PPP, the number of doctors, nurses, and beds per 1,000 inhabitants were determined as the input factors, while amenable mortality was determined as the outcome factor.
The order of the efficiency of the input-output ratio of individual countries was defined using DEA analysis. Subsequently, the countries with the own healthcare financing method were ranked versus the countries with DRG-based financing by the mean value between groups using the Mann-Whitney U-test. The main finding is as follows: there is no statistically significant difference between the mean value of the order of countries with DRG-based healthcare management and financing and countries with the other unique healthcare financing system according to the selected model (p-value is 0.522-0.976 for 2012-2017). Even though DRG-based management and financing have various expected benefits, such as transparency of financing, this approach to healthcare financing and management does not significantly impact amenable mortality.
If any country in the European Union decides to implement DRG-based management and financing management, there can be expected that this change will not significantly influence (improve or worsen) the efficiency set by the selected model of selected inputs and outputs.
The research has some limitations, which are connected with the use of the secondary data from the selected countries of the European Union with generally similar healthcare financing systems. For further