Impact of hemodialysis center accreditation on patient mortality: Korean nationwide cohort study
Article information
Abstract
Background
Since hemodialysis (HD) patients are prone to various complications and high mortality, they need to be treated in HD units with professional personnel, proper equipment, and facilities. The Korean Society of Nephrology has been conducting an HD unit accreditation program since 2016. This study was performed to evaluate whether a qualified dialysis center (QDC) reduced the mortality of HD patients.
Methods
This longitudinal, observational cohort study included 31,227 HD from 832 facilities. HD units were classified into two groups: the hospitals that have been certified as QDC between 2016 and 2018 (n = 219) and hospitals that have never been certified as QDC (non-QDC, n = 613). Baseline characteristics and patient mortality were compared between QDC vs. non-QDC groups using Korean HD quality assessment data from 2018. Multivariate logistic regression and the Cox proportional hazards model were used to compare patient mortality between the two groups.
Results
Among study subjects, 30.6% of patients were treated at QDC and 69.4% were treated at non-QDC. The patients in the QDC were younger and had a longer dialysis duration, lower serum phosphorus and calcium levels, and higher hemoglobin and single-pool Kt/V levels compared to the patients from the non-QDC group. After adjusting for demographic and clinical parameters, QCD independently reduced mortality risk (hazard ratio, 0.897; 95% confidence interval, 0.847–0.950; p < 0.001).
Conclusion
The HD unit accreditation program may reduce the risk of death among patients undergoing HD.
Introduction
The prevalence of end-stage kidney disease (ESKD) is increasing worldwide every year, and healthcare expenditures for dialysis treatment continue to expand [1]. Most ESKD patients already have multiple comorbidities before they start dialysis, including heart disease, cerebrovascular disease, diabetes mellitus, and hypertension [2]. Even after starting dialysis, complications can occur in almost all organs, including the heart, lungs, brain, blood vessels, and digestive system. Because the morbidity and mortality rates of ESKD patients are higher than those of the general population, it is very important to have professional personnel as well as proper facilities and equipment in hemodialysis (HD) units [3,4].
However, despite the rapid increase in the number of ESKD patients and subsequent increase in the number of HD centers, some HD centers still lack sufficient personnel and facilities, threatening the health of patients in Korea [5,6]. In response, the Korean Society of Nephrology (KSN) introduced the HD unit accreditation program in 2016 to recommend and evaluate standard practice guidelines for the safe treatment of HD patients and to manage the quality of HD facilities [7].
A few hundreds of hospitals received their initial qualified dialysis center (QDC) certification in the past few years. Since several years have passed after the HD unit accreditation program was launched, questions have been raised about whether QDCs have better outcomes than non-QDCs and, more importantly, whether the action toward QDC certification improves the outcome of HD treatment. There have been papers on the effect of certification program on patient prognosis in other diseases such as stroke center certification [8], but no research has been conducted on the impact of HD unit certification on patient mortality. To address this point, we compared all-cause mortality of the patients in HD centers who received QDC certification between 2016 and 2018 with those in HD centers without QDC certification (non-QDCs).
Methods
Hemodialysis unit accreditation
In Korea, HD unit accreditation programs have been conducted since 2016 [7]. The hospitals eligible for the evaluation were HD units operated by KSN members that had performed HD for a year or longer. HD units that wish to participate in the accreditation program must enter data on the HD unit accreditation website (http://ksn.nephline.com), and approval is determined through a process of 1) application for certification, 2) online screening, 3) on-site evaluation, 4) comprehensive evaluation, 5) final evaluation, and 6) issuance of a certificate [7]. Once approved, the certification is valid for 3 years.
The four evaluation items used in the HD unit accreditation program include structure, process, ethics, and medical records/reports [7]. The first item, structure, includes the healthcare personnel, facility and equipment, and water treatment system. The staffing requirements were assessed based on the percentage of dialysis specialists, the average number of HD cases per day for physicians, the percentage of nurses with more than 2 years of HD experience, and the average number of HD cases per day for nurses. The second item, process, includes regular HD adequacy monitoring and regular laboratory testing. The third item, ethics, assesses whether there are any unethical practices including solicitation. Lastly, the fourth evaluation item, medical records/reports, includes participation in the ESKD registry program and documentation of death and transfer reports. All four conditions must be met for HD unit accreditation.
Study population and data collection
This longitudinal observational cohort study included Korean patients undergoing maintenance HD. HD units were classified into two groups according to HD unit accreditation: the QDC group (the hospitals that have been certified between 2016 and 2018, n = 219) and the non-QDC group (the hospitals that have never been certified as QDC, n = 613). We have checked the registration number of medical institutions from the QDC group. Then, the registration number of medical institutions from the QDC group was provided to the Health Insurance Review and Assessment Service (HIRA) to match their HD quality assessment data from 2018. The HIRA provided the final matched datasheet with individual patient data according to HD unit accreditation. Patients aged ≥18 years who received HD treatment more than twice weekly were enrolled. Patients who were admitted to the hospital or not followed up during the assessment period were excluded from the HD quality assessment.
HD quality assessment was based on the data entered from each HD facility using a web-based data collection system from March to August 2018 [6]. Demographic and clinical data were obtained from the HIRA database. Demographic factors included age, sex, dialysis duration, body mass index (BMI), and health insurance status were collected. The medical comorbidities of the participants were identified using the International Classification of Disease 10th Revision codes from the health insurance claims database from September 2017 to August 2018. The categories of comorbidities were diabetes mellitus (E10–14), hypertension (I10–13 and I15), ischemic heart disease (I20–25), congestive heart failure (I50), and cerebrovascular disease (I60–64 and I69). Laboratory parameters including plasma hemoglobin (Hb), serum albumin, calcium, and phosphorus levels were collected monthly during the study period. The average values of three measurements were used in the analysis. Dialysis adequacy was evaluated using single-pool Kt/V.
Outcome measures
The primary endpoint of the study was defined as all-cause mortality. The date of death was defined as the date of insurance loss. Patients who received a kidney transplant during the follow-up period were censored at the time of kidney transplant. Mortality data were collected between September 2018 and November 2021.
Statistical analyses
Patient characteristics, hospital characteristics, and all-cause mortality were compared between the QDC vs. non-QDC groups. Baseline characteristics were compared using the chi-square tests for categorical variables and the independent t tests for continuous variables. Continuous variables were expressed as means and standard deviations, and categorical variables as frequencies and percentages. The Kaplan-Meier analysis and the log-rank test were used to compare the risk of death between groups. Multivariate Cox proportional hazard models were used to determine whether QDC groups is an independent predictor of mortality. Model 1 was adjusted for age, male sex, dialysis duration, and BMI. Model 2 was adjusted for medical comorbidities (diabetes mellitus, hypertension, ischemic heart disease, heart failure, cerebrovascular accident, atrial fibrillation) in addition to the factors included in model 1. Model 3 was adjusted for the laboratory parameters (plasma Hb, serum albumin, calcium, phosphorus, single-pool Kt/V) in addition to the factors included in model 2. Finally, subgroup analyses were performed to define mortality risks according to age group (<65 years vs. ≥65 years), sex, dialysis duration (<5 years vs. ≥5 years), comorbidities, plasma Hb (<10.0 g/dL vs. ≥10.0 g/dL), and albumin level (<3.5 g/dL vs. ≥3.5 g/dL). All statistical analyses were performed using R version 4.0.2 (R Foundation for Statistical Computing, http://www.r-project.org/). Standardized mean differences were calculated, and p < 0.05 was considered statistically significant.
Ethics statement
This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of Hallym University Kangnam Sacred Heart Hospital (No. HKS 2021-11-043). The IRB waived the need for written informed consent from the patients because the study participants were de-identified.
Results
Baseline characteristics of participants according to hemodialysis unit accreditation
This study included 31,227 patients treated at 832 facilities. The baseline characteristics of the study subjects according to QDC groups are shown in Table 1. The mean age of total subjects was 60.9 ± 12.8 years. Among them, 60.0% were male, 60.3% had diabetes mellitus, and the mean Kt/V was 1.58 ± 0.28. A total of 9,554 patients (30.6%) were treated at QDC while 21,673 patients (69.4%) were treated at non-QDC. The patients in the QDC group were younger (60.4 ± 13.0 years vs. 61.2 ± 12.7 years old, p < 0.001) and had a longer dialysis duration (6.21 ± 5.59 years vs. 5.95 ± 5.24 years, p < 0.001) compared to the non-QDC group. The subjects in the QDC group showed a lower proportion of diabetes mellitus and ischemic heart disease but a higher proportion of heart failure than the non-QDC group. The QDC group showed lower levels of serum calcium (8.91 ± 0.69 mg/dL vs. 8.93 ± 0.73 mg/dL, p = 0.03) and phosphorus (4.95 ± 1.19 mg/dL vs. 5.04 ± 1.26 mg/dL, p < 0.001) compared to the non-QDC group. Single-pool Kt/V (1.61 ± 0.28 vs. 1.57 ± 0.27, p < 0.001) and plasma Hb level (10.65 ± 0.73 mg/dL vs. 10.63 ± 0.74 mg/dL, p = 0.004) were higher in the QDC group than in the non-QDC group.
Baseline characteristics of the subjects according to HD unit accreditation
All-cause mortality according to hemodialysis unit accreditation
A total of 5,848 deaths (18.7%) occurred within 35.6 ± 8.5 months. After censoring 3,515 cases (10.1%) who received kidney transplantation during follow-up, the crude death rate was 63.2 patients per 1,000 person-years. The QDC group exhibited a lower risk of death than the non-QDC group, showing a crude death rate ratio of 0.86 (Table 2). Kaplan-Meier survival analysis also showed that the patients from the QDC group had better survival during follow-up (log-rank test p < 0.001) (Fig. 1).
Comparison of crude death rate according to HD unit accreditation
Kaplan-Meier survival analysis.
A total of 5,848 deaths occurred during 35.6 ± 8.5 months. After censoring 3,515 cases (10.1%) who received kidney transplantation during follow-up, the qualified dialysis center (QDC) group showed better survival compared to the non-QDC group (p < 0.001).
The Cox proportional hazards model was used to identify the risk factors associated with patient mortality. In the univariate analysis, old age, male sex, lower BMI, presence of comorbidities (diabetes mellitus, hypertension, ischemic heart disease, heart failure, cerebrovascular accident, and atrial fibrillation), lower plasma Hb, serum albumin, calcium, and phosphorus levels were associated with a higher mortality risk (Table 3). In addition, the QDC group was associated with lower mortality (hazard ratio [HR], 0.859; 95% confidence interval [CI], 0.811–0.910; p < 0.001). When we adjusted for age, sex, dialysis vintage, and BMI (model 1), the QDC group was still an independent predictor for patient mortality (HR, 0.89; 95% CI, 0.84–0.94; p < 0.001). When we adjusted for comorbid conditions in addition to factors included in model 1 (model 2), the QDC group remained an independent risk factor for patient mortality (HR, 0.90; 95% CI, 0.85–0.95; p < 0.001). Finally, when we adjusted for laboratory parameters in addition to factors included in model 2 (model 3), the QDC group was an independent risk factor for all-cause mortality (HR, 0.90; 95% CI, 0.85–0.95; p < 0.001).
Multivariate Cox hazards model for patient mortality
Subgroup analysis for all-cause mortality
To define the subgroup who are likely to get survival benefits from HD unit accreditation, we performed a subgroup analysis. The QDC group had higher patient survival in all subgroups except for the elderly (≥65 years), those with cerebrovascular accidents, or Medicaid patients (Fig. 2). The younger patients with a relatively short duration of dialysis were more likely to benefit from HD treatment from QDC.
Forest plot depicting the survival benefit according to hemodialysis unit accreditation in different subgroups.
The qualified dialysis center group had higher patient survival rates in all subgroups except for the patients with old age (≥65 years), cerebrovascular accident, or those on Medicaid.
CI, confidence interval; HR, hazard ratio.
Discussion
This observational study showed that maintenance HD patients in QDC have a 10% lower mortality risk compared to the non-QDC group. The patients in QDC showed lower serum calcium and phosphorus levels, and higher Hb and single-pool Kt/V levels compared to the non-QDC group. However, when we adjusted covariates, the QDC group remained an independent factor for survival.
Many countries have their own approval and/or accreditation programs for HD units [3]. Approval involves the evaluation of several criteria before the establishment of an HD unit. On the other hand, an accreditation program involves the evaluation of maintaining professional personnel, proper facilities, equipment, and services after the establishment of HD units. In the United States, the Survey and Certification Program certifies HD facilities for inclusion in the Medicare Program by evaluating whether each facility meets specified safety and quality standards [9]. The United States government also delegates its role to the ESRD Network, an autonomous organization for quality control and monitoring, to ensure that these facilities continue to meet these basic requirements [10]. Korea, like Japan, does not have an obligatory program for approval or accreditation operated by the Korean government. The lack of standards for approval or accreditation systems may threaten patient health. Although Korean HIRA collects data from all HD units periodically for the evaluation of HD adequacy, it is insufficient to investigate the real-world practice in HD units. For example, some HD units may provide low-quality HD service due to the heavy workload of staff or even provide unethical solicitation to recruit more patients [5,11,12]. Instead, KSN has set up the accreditation program and voluntarily has operated the program since 2016. The evaluation criteria for the accreditation program not only include structure (staffing, facility, and equipment) and HD process (HD adequacy test and regular inspection) but also include ethical aspects (solicitation to recruit patients or providing free transportation) and official records of patient events (death and transfer) [3,7]. The HD unit accreditation program, however, was set up to guide the minimum requirement for HD treatment. Therefore, the program collects whether or not the institution passes certain goals to operate adequate HD treatment.
This study is important to understand whether the self-regulated accreditation program by KSN improves clinical outcomes and reduces mortality in HD patients. Our study suggested that the patients treated in the QDC may have better patient survival and clinical outcomes. Individual characteristics such as age, comorbidities, and nutritional status can be important factors that determine clinical outcomes in HD patients [13]. A previous study by Owen et al. [14] showed that low serum albumin concentration is an important predictor of patient death showing an odds ratio of 1.48 in those with serum albumin concentration of 3.5 to 3.9 g/dL and 3.13 for those with serum albumin concentration of 3.0 to 3.4 g/dL compared to those with normal serum albumin concentration over 4.0 g/dL. Serum albumin concentration was a stronger risk factor compared to the urea reduction ratio. Diabetic patients had lower patient survival due to lower serum albumin concentration and lower urea reduction ratio. The other study performed by Combe et al. [15] also showed that patient survival was significantly influenced by age, the presence of diabetes mellitus, and serum albumin concentration, but not by other variables, including Kt/V and urea reduction ratio. Our study also demonstrated that age and serum albumin concentration affect patient mortality but not Kt/V. However, a recent study by Ajmal et al. [16] demonstrated that facility-level quality control is also an important factor for patient mortality among incident HD patients. In their study, the dialysis facilities with lower Quality Incentive Program (QIP) performance scores, which may be compared to our non-QDC group in the HD unit accreditation program, showed higher patient mortality. The patients with QIP scores <45 (HR, 1.39) and 45 to <60 (HR, 1.21) had higher patient mortality rates than those in the facilities with scores ≥90. The patients with lower QIP scores did not demonstrate differences in age and sex but showed significant differences in BMI and underlying comorbidities (diabetes mellitus, ischemic heart disease, heart failure, and cerebrovascular accident). Our study also showed significantly higher BMI and higher prevalence of comorbidities in the patients in the non-QDC group. Our study also demonstrated lower serum calcium and phosphorus and higher plasma Hb and single-pool Kt/V in the QDC group. These findings can be interpreted as individual characteristics. On the other hand, they can be the results of facility-level quality control among QDC. Our previous study demonstrated that structural and procedural indicators are important for the survival of HD patients [6]. Specifically, a high nurse caseload was associated with high patient mortality in HD units [11]. Dialysis specialist care was also an important determinant of overall patient survival [17]. In addition to adequate staffing, the environmental impact of HD can be essential [18]. Monitoring and maintaining laboratory parameters in adequate ranges can also be essential [16,19]. Although we did not analyze the effect of each component in the accreditation program upon clinical outcomes, HD facilities in the QDC group are likely to make an effort to improve structural and procedural dimensions. In addition to facility-level quality control, ethical components can be important for patient outcomes. Previous report showed that the overall mortality rate in incident HD patients was higher in soliciting facilities compared to non-soliciting facilities [5]. The soliciting facilities are likely to reduce costs by minimizing the dialysis time or using cheaper and less effective equipment or medication.
It is true that baseline profiles of the HD patients were better for patient survival: younger age, lower proportion of diabetes mellitus, better laboratory parameters, and HD adequacy. However, when we adjusted them in multivariate Cox regression analysis, the QDC group was an independent predictor for patient survival. While individual characteristics are not modifiable, we may improve each component in the accreditation evaluation by efforts. Further prospective interventional studies are needed to confirm whether facility-level accreditation improves patient outcomes.
The elderly patients (≥65 years), those with cerebrovascular accidents, or those on Medicaid insurance were unlikely to get benefits from QDC. It is likely that age and history of cerebrovascular accident are stronger risk factors for mortality in the elderly population. The patients on Medicaid insurance were small population and our result cannot be generalized for the Medicaid population. The characteristics of individual patients may have a greater impact on clinical outcomes than the status of accreditation in these subgroups.
There are several limitations in our study. The follow-up duration for mortality analysis is only 35.6 months. Since the HD unit accreditation program launched in 2016, the long-term benefit of the accreditation program should be evaluated in the future. In addition, the effect of each indicator in the accreditation program on patient mortality was not evaluated further. Whether improvement in each indicator results in improved patient survival should be elucidated in the future study. Finally, the certified institutions are more likely to participate in HD unit accreditation programs and therefore there may be a selection bias. The gap between the qualified centers and non-qualified centers can be larger than the presented data. However, we have utilized a multivariate Cox hazards model to adjust possible confounders.
However, this is the first study to evaluate the clinical utility of the HD unit accreditation program in Korea. The accreditation or certification programs on specific diseases or procedures may motivate physicians to improve standard care which may improve patient outcomes [8]. In order to implement basic standards of care in HD units, there should be continuous, cooperative efforts from government body, KSN, and related organizations.
Notes
Conflicts of interest
All authors have no conflicts of interest to declare.
Funding
This study was supported by a grant from the Korean Nephrology Research Foundation (Baxter 2023).
Data sharing statement
The data that support the findings of this study are available from Health Insurance Review and Assessment Service (HIRA), but restrictions apply to the availability of these data, so they are not publicly available. Data are, however, available from the authors upon request and with permission from HIRA.
Authors’ contributions
Conceptualization: DHK, WMH, DJK, YKL
Data curation: BYK, ML, GOK
Formal analysis: DHK, AJC, JK
Funding acquisition: YKL
Investigation: DHK, AJC, YKL
Funding acquisition: YKL
Visualization: HCP, DHK, JK
Writing–original draft: HCP, DHK
Writing–review & editing: AJC, BYK, ML, GOK, WMH, JK, DJK, YKL
All authors read and approved the final manuscript.
Acknowledgements
The authors participated in the Joint Project on Quality Assessment Research, and Health Insurance Review and Assessment Service collected and provided the claims data and quality assessment data to the authors.