Diagnosis related group migration as an indicator of a potentially inefficient medical intervention

Introduction: Diagnosis related group migration (DRG) is a phenomenon in coding hospitalizations that indicates that a patient has been transferred from one group to a different, more costly group, after an episode of hospital treatment, solely because of complications or comorbidities that occurred during hospital treatment or medical intervention.

Objective: The objective of the study was to examine the following: occurrence of DRG migrations, hospital costs, and length of hospitalization for colectomies.

Methods: A retrospective secondary analysis of DRG data was conducted for a series of episodes of hospital treatment (n = 4,939) in patients aged 18 and above, who underwent colectomy in 49 hospitals in the Republic of Serbia, in the period between 2018 and 2020, and whose episodes of hospital treatment were grouped into three diagnosis related groups: the G02A group, a separate G02A group with T81-T88 diagnoses for surgical treatment complications, and the G02B group.

Results: In the observed three-year period, the incidence of DRG migration was 7.4%. Inpatient treatment of patients classified in the diagnostic group G02A (T81-T88), which represents DRG migration, was statistically significantly more costly and these patients had a statistically significantly longer hospital stay. The total invoice amount for the diagnostic group G02A (T81-T88) was 509,651.7 dinars (approximately 4,320 euros), and the average length of an episode of hospital treatment was 23 days. In group G02B, the total invoice amount was 231,989.0 dinars (approximately 1,960 euros) with an average length of hospital stay of 11 days.

Conclusion: Colectomies with complications prolong the length of hospitalization and imply a higher consumption of resources in the Republic of Serbia. Cost analysis related to DRG migration, per patient day and patient, may indicate the ineffectiveness of medical interventions.

INTRODUCTION

The system of diagnosis related groups (DRG) is a modern system for classifying patients treated in hospital (inpatients) into groups which describe different combinations of cases in patients – casemix. This system is being used all over the world with the aim of improving financing models, resource control, and hospital services payment. Casemix is an international term describing the consistent methodology for classifying patients, episodes of hospital treatment (patient episodes), and overall costs (consumption), on the basis of the following: 1) the consumption of hospital resources for patient treatment and 2) clinical manifestations of a disease in a patient. Each episode of hospital treatment, i.e., each patient, is classified into one of the diagnosis related groups. All health services, drugs, and materials spent during a single episode of hospital treatment of one patient are integrated within each DRG [1], in such a way as to enable the analysis of hospital costs by DRGs. Hospitals implementing payment on the basis of the DRG system, may be stimulated to reduce costs by shortening hospital stay, redirecting patient treatment to day hospitals, and reducing the number of diagnostic and treatment procedures [2]. On the other hand, the application of the DRG system carries the risk of hospitals becoming excessively oriented towards cutting costs, which consequently means premature patient discharge and saving on medication and service packages, thus jeopardizing the quality of treatment [2]. Therefore, analyzing DRG migrations is important for a number of reasons. For instance, at the individual level, it helps in clarifying the hospital treatment process of each patient; at the level of the hospital, it contributes to detailed monitoring of the hospital’s operation and management; at the level of hospital health care, it facilitates the comparison of the mode of operation between hospitals; and at the financial level, it may stimulate an equitable distribution of funds amongst hospitals, as well as internal control of treatment costs, and, especially, external monitoring and control of costs.

The methodology of the system of DRG grouping is highly sophisticated and depends on the quality of the data that is used. This is why the quality of hospital medical records is of vital importance in showing whether the degree of resource consumption in episodes of surgical and of conservative treatment corelate with the severity of patient presentation, i.e., with the level of complexity of the comorbidities and/or complications. This information is obtained from the four alphanumeric characters comprising the code for each DRG (Figure 1).

Figure 1. Symbols of the DRG G02A – Major small and large bowel procedures, major complexity and comorbidity

Each DRG, in keeping with the corresponding category, the partition, and the level of consumption, has a defined cost weight, which is used as the basis for calculating the monetary value of the group (Figure 2).

Figure 2. Cost weight of the G02A group – Major small and large bowel procedures, major complexity and comorbidity; cost weight of the G02B group – Major small and large bowel procedures, intermediate complexity and comorbidity

At hospital admission, the patient is classified into the appropriate diagnosis related group, according to the main diagnosis which was the cause for hospitalization. This is done with the aid of the special function of the DRG grouping software – Grouper, which uses a series of algorithms for processing patient data, the patient’s clinical state of health, and the treatment received by the patient. At the end of the episode of hospital treatment, the Grouper checks data on procedures within the episode, in order to categorize the episode, either in the surgical partition, the invasive nonsurgical partition, or the conservative partition. The Grouper checks the age and sex of the patient, the underlying diagnoses, and the treatment outcome, and it determines the length of the episode of hospital treatment. DRG migration means that the patient, i.e., the episode of hospital treatment becomes categorized, not in the same, but in a different, more costly group, exclusively due to complications occurring during treatment or surgical procedure. This means that treatment has led to a certain level of deterioration of the patient’s health status, as well as to an increase in resource consumption. This is why DRG migration in a healthcare facility (or an individual department) may be used as an indicator of the efficiency and quality of health care being provided. This is especially important for surgical branches of medicine, which are costly, and often require a long period of recovery in hospital, as well as significant funds [4].

In some hospitals, there are employees who deal with DRG coding and control before the records and documentation are handed over to the insurance companies [3]. In the Republic of Serbia, the National Health Insurance Fund of the Republic of Serbia (RFZO) is predominantly the buyer of services provided by the stateowned hospitals. As of January 1, 2019, the RFZO has been applying the payment model based on DRGs, along with line item (dedicated) budgeting – the budget is negotiated based on last year’s contracts, while any remaining funds are returned to the RFZO. At the moment, 95% of funds are defined by estimate and transferred every month to healthcare institutions, by budget lines, while the remaining 5% of funds are the variable part of the budget whose payment depends on DRG performance [5]. DRG performance is assessed with the Australian Refined Diagnosis Related Groups Version 6.0, which also has software for data grouping. According to the assessment of the Second Serbia Health Project, in the period between 2018 and 2020, the amount of approximately 142 billion RSD (approximately 1.2 billion euros) was designated annually for health institutions at the secondary and tertiary levels of health care. Of this sum, more than one third (around 44.8 billion RSD a year) was designated for hospitals, while, on average, 32% of the budget of an individual hospital was allocated to surgical departments, with this percentage running as high as 90% or above, in some hospitals [6].

To the best of our knowledge, DRG migration and the cost analysis of surgical procedures, with and without complications, have as yet not been the subject of scientific research in the Republic of Serbia. In line with the practice of developed countries, it is necessary to scientifically and professionally justify the costs of the most expensive forms of treatment, such as hospital treatment, especially in situations when large investments are planned for inpatient facilities [7]. Currently, the attention of the professional circles as well as of the public is being directed towards determining the difference in costs among colectomies, treatment with laparoscopic procedures, and open surgery of the colon [8],[9], as well as towards monitoring costs of the use of cutting-edge technology and robotics [10]. International publications have demonstrated that DRG migration is used as one of the tools for assessing the quality and efficiency of work, both at the department and hospital levels, as well as at the national level [5].

Colectomy, as a surgical procedure, is used in treating diseases of the colon other than colon cancer. However, bearing in mind that that colon cancer is one of the three most common malignant diseases in the world population, it is clear that this procedure is performed rather frequently. According to available data, colon cancer is second (11.2%) on the list of leading localizations of cancer-related morbidity in the general population of the Republic of Serbia, and it is also second (12%) on the list of leading localizations of malignancy-related mortality [11]. Therefore, the aim of this paper is to investigate the following: occurrence of DRG migration in colectomies, i.e., categorizing the patient episode into the DRG group which is potentially characterized by a higher treatment cost; hospital costs; length of hospital stay.

MATERIALS AND METHODS

A retrospective secondary analysis of data on DRG was performed, for series of episodes of hospital treatment (n = 4,939), in patients aged 18 and above, who had undergone colectomy, in 49 hospitals belonging to the Plan of the Republic of Serbia Healthcare Institutions Network [12],[13], in the period between 2018 and 2020, and whose episodes of hospital treatment were grouped into three DRGs:
1. G02A – Major small and large bowel procedures, major complexity and comorbidity (cost weight = 6.8)
2. G02A (T81-T88) – Major small and large bowel procedures, major complexity and comorbidity (cost weight = 6.8) – separate associated diagnoses, which may represent complications of surgical treatment that were the cause of DRG migration
3. G02B – Major small and large bowel procedures, intermediate complexity and comorbidity (cost weight = 3.51).

The data source used for the study is the database of the Second Serbia Health Project implemented by the Ministry of Health [6], which generates and further analyses all data obtained from public healthcare institutions through electronic billing. For the purpose of this analysis, data for the period between January 1, 2018 and December 31, 2020 were taken from the database.

With respect to colectomies, 12 different surgical procedures have been defined, with associated codes [14]:
1. 32003-00 Limited excision of large intestine with anastomosis
2. 32000-00 Limited excision of large intestine with formation of stoma
3. 32003-01 Right hemicolectomy with anastomosis
4. 32000-01 Right hemicolectomy with formation of stoma
5. 32005-01 Extended right hemicolectomy with anastomosis
6. 32004-01 Extended right hemicolectomy with formation of stoma
7. 32006-00 Left hemicolectomy with anastomosis
8. 32006-01 Left hemicolectomy with formation of stoma
9. 32005-00 Subtotal colectomy with anastomosis
10.32004-00 Subtotal colectomy with formation of stoma
11.32012-00 Total colectomy with ileorectal anastomosis
12.32009-00 Total colectomy with ileostomy.

Guided by scientific studies in this field [8],[9], the following diagnoses, which may present complications of surgical treatment, and which provide the supposition of DRG migration to the group G02A (T81-T88), were taken:
1. T88.8 – Other complications of surgical and medical care, not elsewhere classified
2. T81.0 – Hemorrhage and hematoma complicating a procedure, not elsewhere classified
3. T81.3 – Disruption of operation wound, not elsewhere classified
4. T81.4 – Infection following a procedure, not elsewhere classified
5. T81.5 – Foreign body accidentally left in body cavity or operation wound following a procedure
6. T81.8 – Other complications of procedures, not elsewhere classified
7. T81.9 – Unspecified complication of procedure.

The variables included in DRG (G02A, G02A (T81-T88) and G02B) analysis, are as follows:
1. The predominance of different diagnoses which represent the basic cause of hospitalization (BCH)
2. Patient age, expressed as the mean and as standard deviation (SD)
3. Patient sex
4. Length of hospital stay (number of days)
5. Hospital treatment costs – seven types of costs expressed as the median (25th and 75th percentile), include the following:
• Total value of services rendered (without patient days)
• Total value of billed patient days
• Total cost of medication
• Total cost of blood and blood products
• Total cost of disposable medical supplies
• Total invoice amount.
6. Serbia, in total, and by regions – City of Belgrade, East Serbia, South Serbia, Central Serbia, Autonomous Province of Vojvodina, and West Serbia.

In the observed three-year period, there was a total of 4,939 observed groups, amongst which group G02B was predominant, with a prevalence of 70.5%.

Statistical data analysis

Data were analyzed with the use of descriptive and analytical statistics methods. The following descriptive statistics methods were used: measures of central tendency (mean, median), measures of variability (standard deviation and percentiles), absolute and relative numbers. Normality of distribution was tested with graphic and mathematical methods. For the comparison of statistically significant difference amongst the analyzed groups, the ANOVA test (or the Kruskal-Wallis test) was used, i.e., the t-test (or the Mann-Whitney test), for numeric data, and the χ2 test (or Fisher’s test), for nominal data. Statistical significance was accepted at the level of 0.05. The results have been presented in tables. The SPSS 21.0 software package was used for data processing.

RESULTS

In the period between 2018 and 2020, in 49 hospitals in Serbia, the diagnoses that were the most prevalent as the basic cause of hospitalization, in groups G02A and G02B, were C18.7 – Malignant neoplasm of the sigmoid colon and C18.2 – Malignant neoplasm of the ascending colon, followed by C18.0 – Malignant neoplasm of the appendix (Table 1). In the group G02A (T81-T88), the majority of the patients had, as the basic cause of hospitalization, the following diagnoses: K56.7 – Ileus, unspecified, K56.6 – Other and unspecified intestinal obstruction, and C18.0 – Malignant neoplasm of the appendix.

Table 1. The most prevalent diagnoses as the basic cause of hospitalization and the number of treatment episodes, by DRGs, for major small and large bowel procedures, Serbia, 2018 - 2020

In the observed period, in the diagnostic group G02A, there were, in total, 1,349 patients. Their average age was 69.8 years. At the same time, in group G02B, the number of patients was 2.5 times greater (3,482), and their average age was 64.9 years. A total of 108 (7.4%) patients, in whom some complication was registered upon surgery and whose average age was 70.6 years, migrated from the G02B group to the separate group G02A (T81-T88). In all three groups, a majority of the patients were male. Statistically significant differences were determined in all the observed costs and in the length of hospital stay, in relation to all the observed DRGs (Table 2).

Table 2. The differences in patient characteristics, costs and length of hospital stay, by observed DRGs, Serbia 2018 - 2020

As opposed to the group structure pertaining to the sex of the patients, with regards to the age of the patients, the difference amongst the groups was statistically significant, with the oldest patients belonging to the group G02A (T81-T88) (Table 2). Post hoc analysis determined a statistically significant difference between groups G02A and G02B, as well as between groups G02A (T81-T88) and G02B, but not between groups G02A (T81-T88) and G02A (Table 3).

Table 3. Post hoc analysis of patient characteristics, costs, and length of hospital stay, according to the observed DRGs, Serbia, 2018 – 2020

The total invoice amount for group G02A was 290,702.9 RSD (approximately 2,460 euros), and the average hospital stay was 13 days. The total invoice amount for group G02A (T81-T88) was 509,651.7 RSD (approximately 4,320 euros), while the average episode of hospital treatment was 23 days. The total invoice amount for group G02B was 231,989 RSD (approximately 1,970 euros), with an average hospital stay of 11 days (Table 2).

The value of services rendered without patient days, the total value of patient days, the total cost of medication, as well as the cost of blood and blood products, were the highest in group G02A (T81-T88), and for these patients, hospital stay was the longest. Post hoc analysis has determined that the differences in the values of services rendered without patient days, in the total value of patient days, the total medication costs, as well as in the number of patient days, were statistically significant between group G02A (T81-T88) and group G20A, between groups G02A and G02B, as well as between groups G02A (T81-T88) and G02B (Table 3). The statistical significance of these differences in costs has been confirmed at the level of all the regions (Table 4).

Table 4. The differences in patient characteristics, costs and length of hospital stay, by observed DRGs, Serbia 2018 - 2020

In all the regions, except West Serbia, the total invoice amount was the highest for group G02A (T81-T88) (p < 0.001) and ranged between 390,048 RSD (approximately 3.300 euros), in Central Serbia, to as much as 581,338.8 RSD (approximately 4,930 euros), in the South Serbia region. The longest average hospital stay was registered for group G02A (T81-T88) and ranged from 21 days, in the City of Belgrade, to 31 days, in the region of Central Serbia, while it was the shortest in the region of West Serbia (11 days on average) (p < 0.001) (Table 4).

DISCUSSION

During the observed three-year period, DRG migration was registered in 108 (7.4%) episodes of hospital treatment, which means that patients, due to complications (diagnoses T81-T88) which had developed during hospital stay were categorized into another DRG, with a greater cost weight. However, this data must be considered with certain reservations, since not all complications potentially leading to DRG migration could be identified in the process of data analysis. Namely, the most frequent complications following large intestine excision are different wound infections and infections of internal organs, followed by gastrointestinal motility complications, including ileus and bowel obstruction [14]. However, for the said diagnoses, it cannot be determined with certainty when they developed and in which part of the patient’s hospital stay – before, during or after the surgical procedure, as opposed to the observed diagnoses in the analysis of T81-T88, whose name clearly indicates complications occurring as the result of a surgical procedure or treatment. This problem is the result of the method used for reporting on the functioning of healthcare institutions to the RFZO. Namely, the electronic invoice does state the basic cause of hospitalzation and all the underlying diagnoses that were significant for the episode of hospital treatment, but these are not further classified to show which comorbidities the patient had been admitted with and which complications occurred during hospital treatment.

As opposed to the more developed countries, where one of the performance-related payment models is applied, in the Republic of Serbia, neither the service provider nor the service contractor have completely accepted the significance of more detailed case reports and files whose purpose is to account for the funds spent on treatment. The reason for this is that money is still mostly (95%) allocated to healthcare institutions, by dedicated budget lines, based on funds contracted on the basis of estimate, while the remaining, variable part, amounting to 5%, is calculated based on the sum of DRG cost weights. Only after the base rate is determined can the costs of each DRG group be defined, upon which the implementation of the payment model based on the DRG system can be considered completely finished. When the practice of defining the budget of healthcare institutions on the basis of the value of their DRG groups becomes prevalent, the significance of proper reporting will increase, both for the hospitals, but also for the RFZO, naturally, for opposing reasons. The interest of the healthcare institution is to justify financial investment, while the interest of the RFZO is to rationalize the consumption of resources.

When the distribution of groups G02A, G02A (T81-T88) and G02B is observed, out of a total of 4,939 patient groups recorded in the period between January 1, 2018 and December 31, 2020, 2,779 groups, i.e., 56.26%, were groups of patients treated at 11 out of a total of 49 hospitals, where colectomies are performed. All of the eleven healthcare institutions are at the tertiary level of health care, which is to be expected, bearing in mind the complexity of the procedures themselves as well as the severity of patient presentation requiring colectomy. With regards to DRG migration, a third (33 out of 108) of the G02A (T81-T88) groups belonged to the abovementioned tertiary healthcare institutions, and DRG migration was only 1.58%. As DRG migration can be considered a measure of the quality of work of a healthcare facility, the tertiary-level institutions with the best medical equipment and the best trained subspecialists also have the lowest DRG migration [15],[16].

International practice has shown that in, so called, open surgery procedures the risk of complications is higher, and thereby the recovery period for the patient is longer, i.e. the risk of prolonged hospital stay is higher. Transitioning to minimally invasive, i.e., laparoscopic colectomies would, in this case, have a positive effect on both outcomes and cost reduction [17]. However, what would be most important is performing regular measuring of this indicator of health care efficiency and incorporating econometry into health policy, since measuring costs, which are generated in the system by complications, particularly evidence-based measuring – as shown in this study, is of great significance [18]. Individual studies have shown that certain complications of colectomies may amount to more than 8% of overall hospital costs [19].

Regular monitoring of indicators of hospital costs and work quality enables the detection of those hospitals which are centers of excellence for certain pathologies. Identifying centers of excellence is very important, since patients can then be referred for treatment to those facilities where the volume of a certain pathology is significant – and international experience has shown that, in such cases, the percentage of adverse effects decreases, the quality of health care improves, and treatment costs drop [20]. This suggest the need for decision makers to consider an organizational model of service provision wherein colectomy cases would be more successfully resolved at tertiary healthcare institutions and at several referral centers, which have the best medical equipment and the best trained subspecialists, and therefore lesser DRG migration.

Regarding colectomies, the average expenditure for these procedures, in the period 2018 – 2020, was around 488.8 million RSD (approximately 4.2 million euros), per year. The results obtained in this study show statistically significant differences in invoice amounts. They also show that the costliest invoices were for group G02A (T81-T88). Also, the longest hospital stays are related to this group. When the rate of DRG migration of 7.4% was extrapolated to all the patients in the Republic of Serbia who had undergone colectomy in the observed three-year period, the result was an increase in the total invoice amount of 2.6% (12.5 million RSD, i.e., approximately 106,500 euros), as well as an increase in the number of patient days by 1,771 days [18]. This practically means that there is an increase of around 590 patient days a year, i.e., almost two occupied beds during the entire year resulting from complications from the T81-T88 spectrum, in colectomy procedures alone. This is why monitoring DRG migrations is a very good indicator for detecting inefficient medical procedures.

The percentage of complications during hospital treatment is one of the indicators of the quality of health care, while the monitoring of DRG migrations occurring as the result of complications developing during procedures performed in the course of hospitalization more specifically indicate the work quality in hospitals. In the Republic of Serbia, the control of the quality of health care is defined by a special Rulebook [21]. The complications defined by the diagnoses T81-T88 are primarily different infections related to surgical procedures, which healthcare institutions are obliged to report on, within the indicators of patient safety in surgical wards. According to the latest available Report on the Improvement of Work Quality in Healthcare Institutions of the Republic of Serbia [22], the incidence rate of surgical site infection is around 1%. However, awareness in the institutions on the importance of recording these adverse effects is still lacking, as is the awareness of their importance, not only for the health of the patient, but also for the overall healthcare system costs, as shown in this study.

The advantages of this study are reflected in the fact that it has offered insight into parameters, such as: distribution of patients who had undergone colectomy by groups, average cost and average length of hospital stay for each of the diagnosis related groups, as well as the percentage of DRG migration, based on the separate complications. On the other hand, bearing in mind that the drawbacks in the process of data collection in the Republic of Serbia lead to the generation of incomplete and imprecise data on all the possible complications following large intestine excision (the already mentioned infections of the wound and internal organs, motility complications, ileus, and intestinal obstruction), this study could not encompass and analyze all potential cases of DRG migration.

CONCLUSION

At the national level, and at almost all the regional levels in the Republic of Serbia, the patients classified in the diagnostic group G02A (T81-T88), which represents DRG migration, are statistically significantly older than the others, and their treatment is statistically significantly longer and more costly, with regards to the value of the services provided, the value of the patient days, the cost of the medication administered, the cost of the blood and blood products, as well as with regards to the total invoice amount. Colectomies with complications, resulting from hospital treatment or medical procedures, lead to the categorization of the given episode of hospital treatment in the diagnosis related group that entails a greater consumption of resources and longer hospital stay. DRG migration, as an indicator of efficiency and quality of provided health care may be a valid signal for control and management of the work process and its quality, in a healthcare institution or its organizational unit.