DRGs-based health system performance for breast cancer patients

Background: To evaluate health system performance for patients with breast cancer in Henan Province, China, using Diagnosis-Related Groups (DRGs) indicators and provide data to inform practices and policies for the prevention and control of breast cancer. Methods: The data were collected from the front pages of the medical records (FPMR) of all hospitals above class II that admitted breast cancer patients in Henan Province between 2016 and 2019. Breast cancer patients were the subjects in our study. China DRGs (CN-DRGs) were used as a risk adjustment tool. Three indicators, including the Case-Mix Index (CMI), number of DRGs, and total weight, were used to evaluate the range of available services for patients with breast cancer, while indicators including the Charge Efficiency Index (CEI), Time Efficiency Index (TEI) and inpatient mortality of low-risk group cases (IMLRG) were used to evaluate the medical service efficiency and medical safety. Results: Between 2016 and 2019, there were 103,760 cases of patients with breast cancer. The number of enrolled patients and total weight increased over the study period at an average annual rate of 21.38% and 21.88%, respectively. The TEI decreased over the study period by 15.60%. The CEI exhibited an increasing trend, but the average annual rate of increase was small (2.94%). The IMLRP was 0.02%, 0%, 0% and 0.01% in 2016, 2017, 2018 and 2019, respectively. Conclusion: The health system performance improved between 2016 and 2019 for breast cancer patients discharged from the study hospitals in Henan Province. The main areas of improvement were in the range of available services, but medical institutions must still make efforts to improve the efficiency of medical services and ensure medical safety. DRGs are an effective evaluation tool. Accreditation of Healthcare Organizations ; KPI :Key Performance Indicators; ICD-10 International Classification of Diseases, Tenth Revision; ICD-9-CM-3 :International Classification of Diseases Clinical Modification of 9th Revision Operations and Procedures; BJ-DRGs : Beijing DRGs; MDCs : Major Diagnostic Categories ; ADRGs : Adjacent DRGs.

selection of the evaluation tool itself [5].
Health system performance evaluation tools are diverse and vary worldwide. For example, in Singapore, models based on the Singapore Quality Award criteria and the Balance Score Card (BSC) approach are used to evaluate the performance of hospitals [6]. The BSC method provides a framework that focuses on key management processes and evaluates the realization of the vision and strategy of a hospital based on the following four dimensions: finance, customer service, internal business, and innovation [7]. In America [8], the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) implemented evidence-based standardized measures of performance in more than 3,000 accredited hospitals. The measures were designed to track performance over time and compare hospitals based on the following six dimensions: safety, patient satisfaction, efficiency, clinical quality, financial management, and medical expenses [9]. In China, the primary tools used for performance management in hospitals focus on the following: financial management, human resource management, and clinical management [10]. Different evaluation methods are chosen according to the different evaluation objects, such as hospitals, departments and doctors [11]. For instance, the BSC method, Key Performance Indicators (KPI) and achievement measurement method are used to evaluate the performance of hospitals, departments and doctors, respectively [5]. To achieve an ideal performance evaluation system, the most important feature is the accuracy of the evaluation results [12]. However, due to the inherent characteristics of medical services, including diversity, high risk, and difficulty in comparison, performance evaluations without risk adjustment, such as key performance indicators, etc., cannot guarantee reliable results [13]. Therefore, an evaluation tool based on risk adjustment can improve the accuracy of the evaluation results.
The Diagnosis-Related Groups (DRGs) system was the first health management tool to group patients into clinically meaningful categories representing equivalent health resource usage. The DRGs system was first adopted in the State of New Jersey in 1980 [14] and was implemented by the US federal government as a payment system in 1983 [15]. Subsequently, several countries adopted the DRGs system [16]. Currently, this system is the most widely used risk adjustment tool [17]. Considering the importance of risk adjustment in performance evaluations, researchers at Peking University began to study the DRGs system as a performance evaluation tool since 2005 [12,18,19]. A series of health system performance evaluation indicators was constructed to evaluate the range of available services for patients, service efficiency, and medical safety. The DRGs system has been shown to have several advantages over traditional evaluation methods. First, in contrast with the subjectivity of the scoring system used in the BSC [20], the DRGs evaluation indicators are based on objective data, such as the number of discharged patients, length of stay, medical cost and mortality. Therefore, an evaluation method using DRGs is likely to be more reliable and accurate than other methods. Second, the DRGs system effectively avoid biases in comparisons by adjusting case-mix across different hospitals [21]; thus, the results are more reliable and impartial [22,23]. Third, this evaluation method is nonexclusive and can be combined with other performance evaluation methods [24]. Finally, continuous data are relatively easy to obtain because they are collected from medical records. Due to these advantages, health performance evaluations conducted in Beijing are based on the DRG system [12,18,19]. This line of studies has more recently been extended to other parts of China [25,26], but to date, no study assessing DRGs-based health system performance in breast cancer patients has been performed.

Methods 1 Data sources
Since 2012, medical institutions in Henan Province have adopted a common and uniform discharge abstract, commonly referred to as the front page of the medical records (FPMR). The FPMR contains much information, including patient demographic information (age, gender, address, etc.), date of admission and discharge, diagnosis (principal diagnosis and other diagnoses), procedures (principal procedures and other procedures), hospitalization outcome, medical costs and prescription records. In this study, FPMRs from all hospitals above class II that admitted breast cancer patients from Henan Province between 2016 and 2019 were reviewed. The relevant information of each case was carefully collated and assessed. Cases were included if they met the following criteria: (i) the ICD-10 code of principal diagnosis contained C50: malignant neoplasm of breast, and (ii) the date of discharge was between January 2016 and December 2019. Cases were excluded if they met the following criteria: (i) the length of stay was longer than 60 days and (ii) critical information, such as the patient's age or gender, diagnoses, procedures performed, discharge date, medical costs, length of stay, etc., was missing. Based on these criteria, we collected 103,760 records between January 2016 and December 2019.  [27]. Figure 1 shows the grouping path of the CN-DRGs.

DRGs evaluation indicators
According to the study conducted by Jian W [12,18,19], we constructed six objective health system performance indicators to evaluate the available range of medical services, efficiency, and safety. The average level of the DRG indicators of the hospitals included in the BJ-DRGs evaluation was selected as the standard during the calculation of the DRG indicators (Table 1).

(1)Service availability indicators[19]
The services availability can be evaluated by calculating the number of DRGs, total weight, and the Case-Mix index (CMI). Together, these indicators captured the range of services available, the total output of inpatient services and the technical difficulty in treating patients after adjusting for each hospital case mix.
(2)Service efficiency indicators [12,19]: Service efficiency is captured by the following two indices: Charge Efficiency Index (CEI) and Time Efficiency Index (TEI). Both indicators represent relative values that can be used to capture the cost and length of stay of treating similar diseases. The larger the two values, the lower the health service efficiency. If the TEI and CEI are greater than 1, the time efficiency and cost efficiency required to treat the same disease are lower than those in the standard sample.
(3) Medical safety indicators [12,18,19]: Inpatient mortality of low-risk group cases (IMLRP) represents the mortality rate from diseases that are extremely unlikely to cause death. A higher mortality rate is indicative of potential clinical process errors. The procedure used to assign patients to lower risks was as follows: (1) The mortality rate (Mi) of each DRG was calculated; (2) the natural logarithm of these mortality rates (Ln(Mi)) was computed following a normal distribution, and the mean and standard deviation of Ln(Mi) were derived; and (3) IMLRP referred to the mortality rate of those DRGs in which Ln(Mi) was less than one standard deviation below the mean value of Ln(Mi).

Statistical analysis
The continuous variables were expressed as the mean ± standard deviation. Ratios or rates were used to describe the categorical variables. Line charts were adopted to describe the time trend of the continuous variables. All analyses were conducted in SPSS version 22.0.     We adopted the CN-DRGs (2018 edition) grouping method to analyze the health system performance in patients with breast cancer. This analysis showed that the health system performance improved during the study period, particularly in the total number of weighted cases as represented by total weight and the length of stay of those cases as represented by the TEI. The total number of weighted cases increased yearly, and the CMI value increased marginally. The number and total weight of the enrolled cases of the six DRGs (JR15, JA15, JB15, JR11, JA13, and JA25) also increased yearly. JR11

Results
(breast malignancy with severe complications or comorbidities) and JA13 (total mastectomy with general complications or comorbidities) were the two fastest-growing groups. This finding suggests that the medical service availability in Henan Province improved continuously throughout the study period. Three reasons could explain this improvement.. First, the introduction of policies, such as the regional medical union, "County-level Clinical Key Specialty Project" and "the 515 Action Plan", increased financial support for medical institutions and improved access to primary health care institutions. Second, the construction of Specialist League,such as Critical Care Medicine Specialist League in Henan Province, help each other and make progress together [29].Third, The development of telemedicine has promoted the construction of the hierarchical medical system [30]. The remote center of one Third-class hospital in Henan province collected more than 70,000 cases from 2015 to the expansion of the adoption of medical teamss [32].
The TEI of all enrolled breast cancer patients was greater than 1 between 2016 and 2019 but decreased yearly, with the largest decrease observed in JA25 (subtotal mastectomy without complications or comorbidities). This finding indicated that the time efficiency of treating similar diseases in medical institutions in Henan Province was worse than that in institutions in Beijing, but it had improved yearly, particularly in patients who underwent surgery. This finding may be associated with the implementation of clinical pathways [33,34]. The CEI of all enrolled breast cancer patients

Conclusions
DRGs represent a risk adjustment tool that can be used to compare the performance of a health system across different years more impartially. Between 2016 and 2019, the total number of weighted cases as represented by total weight increased from year to year, indicating that the medical service availability in Henan Province continuously improved. The TEI was greater than 1 but decreased yearly, while the CEI was less than 1 but slightly increased between 2016 and 2019, indicating that the medical efficiency had improved but needs greater improvement. Two deaths occurred among the low-risk group cases, illustrating that there are opportunities for greater improvement in medical safety. Number of enrolled cases of major DRGs varies monthly Total number of weight cases of major DRGs varies monthly Charge efficiency index (CEI) of major DRGs varies monthly