Outcome and Risk Factors for Mortality in Peritoneal Dialysis Patients: 22 Years of Experience in a Turkish Center

Background: European peritoneal dialysis populations have identified and reported mortality and morbidity risk factors. However, no reports are pointing out the factors affecting the outcomes of these patients during more than 2 decades of follow-up in Türkiye. This single-center study aims to evaluate patient mortality and peritonitis rates and estimate con - founding factors affecting patient mortality over 22 years. Methods: Adult patients who underwent peritoneal dialysis at our center between December 1994 and December 2016 were enrolled in this retrospective cohort study. The primary outcome of the present study was mortality, and the secondary outcomes were technical failure and peritonitis. Results: Two hundred fifty patients were included in this study. The patients were followed up for a median of 39.5 months (range 17-71). Forty-eight (19.2%) patients died. Survival rates at 5, 10, and 15 years were 86.8% (217/250), 64.6% (86/133), and 41.1% (30/73), respectively. The prevalence of diabetes mellitus [14 (29.2%) vs. 20 (9.9%); P < .001] and cardiovascular disease [16 (33.3%) vs. 24 (11.9%); P < 0.001] were significantly higher in the deceased group compared to the survival group. Cardiovascular disease was the leading cause of death [26 (54.1%)]. Age (hazard ratio (HR) 1.06; 95% CI, 1.04-1.09; P < .001), male sex (HR 2.07; 95% CI, 1.10-3.90; P = .024), and transfer to peritoneal dialysis due to vascular access problems (HR 3.91; 95% CI, 1.90-8.07; P < .001) were associated with mortality in multivariate analysis. Also, catheter exit-site infec - tion, peritonitis rate, catheter removal, and technical complications were similar between the groups. The peritonitis rate was 0.2 episodes per patient per year. Conclusion: The mortality rate of the patient population in our center was similar to Europe and the United States. Cardiovascular diseases and diabetes are the leading causes of death in Turkish peritoneal dialysis patients, as in other populations.


INTRODUCTION
Kidney replacement therapies are essential for the end-stage kidney disease (ESKD) patient's survival.Hemodialysis (HD) is the leading kidney replacement treatment, but more than 150,000 patients are under peritoneal dialysis (PD) worldwide. 1According to the Turkish Society of Nephrology, the prevalence of PD was reported as 4.1%. 2 Several conditions, including uremic toxins, peritoneal catheters, and dialysis solutions, may prompt cardiovascular disease in PD patients. 3,4Also, demographic data (e.g., older patients, body mass index (BMI) <18 kg/m 2 ) and comorbid diseases (e.g., cardiovascular diseases and diabetes mellitus) have been documented as risk factors for mortality in these patients.In addition, several European PD populations have identified and reported risk factors for mortality and morbidity. 5However, there were no reports in Türkiye that pointed to factors influencing the outcome of these patients during more than 2 decades of follow-up.This single-center study aims to evaluate patient mortality and peritonitis rates and estimate confounding factors affecting patient mortality over twenty-two years.

Patient Characteristics
Adult patients undergoing PD at our center between December 1994 and December 2016 were enrolled in this retrospective cohort study.Patients with more than 3 months of clinical follow-up under PD were included in the analysis.Patients lost to follow-up were excluded (Figure 1).
The demographic, clinical, and laboratory results were collected from the hospital data and analyzed retrospectively.Age at the initiation of PD, gender, primary kidney disease, history of kidney replacement therapy, comorbidity, previous peritonitis episodes, Kt/V, and creatinine clearance values were obtained.Coronary artery disease, cerebrovascular disease, and peripheral arterial disease are defined as cardiovascular diseases.Adequacy of dialysis was assessed annually through measurement of weekly Kt/V for urea and weekly creatinine clearance using standard methods.Mean values of Kt/V urea and creatinine clearance during follow-up were recorded for each patient.The study obtained approval from the Medical Ethics Committee of Istanbul University Faculty of Medicine on February 26, 2016, with protocol number 2016/269.Informed consent was obtained from the study participants.

Study Outcomes
The primary outcome of the present study was mortality during follow-up in PD patients.The secondary outcomes were PD technical failure and peritonitis in PD patients.

Statistical Analysis
Patients were categorized as survivors and deceased.All variables were analyzed according to the groups.Gender, etiology of chronic kidney disease, comorbid diseases, kidney replacement history, complications, and reasons for PD choice were presented as numbers and percentages.Body mass index, weight, follow-up period, peritoneal creatinine, urea clearance, and residual clearance (with nonnormal distribution) were specified with medians and interquartile ranges.Chi-square and Fisher's exact tests were used for categorical variables, and the Mann-Whitney U-test was used for quantitative variables with a non-normal distribution.
Cox regression analysis was used to determine potential confounders and mortality.Known risk factors (age, sex, cardiovascular diseases, diabetes mellitus) and significant variables between groups (reason for PD choice, prior kidney replacement history) were entered into the univariate analysis.A multivariate analysis was calculated by using confounders obtained from univariate analysis.Any collinearity was not detected, and effect modifiers were not added in the Cox regression analysis.Hazard ratios and the corresponding 95% CIs refer to the increase per unit in a continuous variable.A P-value of less than .05 is considered significant.

MAIN POINTS
• The present study analyzed 22 years of experience with 250 peritoneal dialysis patients with a follow-up of 4 years at a large university hospital.The mortality rate of the peritoneal dialysis population at our center was similar to Europe and the United States.• Patients transferred to peritoneal dialysis due to vascular access problems have a higher mortality rate than those who choose it themselves.Choosing peritoneal dialysis before vascular access problems in hemodialysis patients may reduce mortality rates.• There were no differences between the survival and nonsurvival groups in body mass index and first-, fifth-, and tenth-year weight differences.• The peritonitis rate was 0.2 episodes per patient per year.
The peritonitis rate of our center is below the target of the International Society of Peritoneal Dialysis.Patients were followed with a median duration of 39.5 months (interquartile range (IQR) 25-75, 17-71).One Hundred and eleven patients (44.4%) had started PD first, and 116 (46.4%) had switched from HD to PD.Also, the number of initial PD patients in the surviving group was significantly higher than the deceased patients [98 (48.5%) vs. 13 (27.1%);P = .007],and the number of patients were transferred from HD to PD was lower in the surviving group compared to the other group [86 (42.6%) vs. 30 (62.5%); P = .013)(Figure 2).There were no differences between the surviving and deceased groups in BMI, first-, fifth-, and tenth-year weight differences.
During the follow-up period, 48 patients died.4.

DISCUSSION
The present study analyzed 22 years of experience of 250 PD patients at a large university hospital with a median follow-up  of 39.5 months.The 5, 10, and 15-year survival rates were 86.8% (217/250), 64.6% (86/133), and 41.1% (30/73), respectively.Also, kidney transplantation provides a survival benefit compared to dialysis in initial replacement therapy.In a previous study from Türkiye, the fifth and tenth year survival rates were described as 68.8% and 40.7%, respectively. 6A previous report from ANZDATA found that 3-year patient survival was 65%-73%. 7Our survival rates are better than these reports; however, ANZDATA data was obtained from an older database, and the developments in PD treatment can explain this condition (Table 5).Some articles have reported that high peritoneal membrane permeability is associated with an increased risk of death in the patient population due to increased protein loss with decreased fluid and minor solute removal. 8On the other hand, this argument has not been proven yet. 9In our data set, patients with high peritoneal membrane permeability had a significantly higher survival rate than patients with low peritoneal membrane permeability.However, these results should be confirmed by multicenter studies with a large cohort since the groups were not homogeneous in our single-center study.
According to the US and European Registry Reports, the most common cause of ESKD in PD patients is diabetes mellitus, accounting for approximately 30% of all etiologies. 10,11In our study, chronic glomerulonephritis was the most common cause of ESKD, followed by congenital anomalies of the kidney and urinary tract and diabetes mellitus.However, patients with diabetic nephropathy had a worse prognosis than nondiabetic patients, similar to previous registry reports.
In our study, the frequency of cardiovascular disease and diabetes mellitus was higher in the deceased group than in the survived group.However, multiple regression analyses showed that age and transfer to PD due to vascular access problems were associated with mortality.This result was associated with the relatively low number of diabetic and cardiovascular patients.
The peritonitis rate of our center is below the International Society of Peritoneal Dialysis 12 target of peritonitis rate at risk, which does not exceed 0.4 episodes per year.Our study found that a history of peritonitis was a predictor of mortality.The reasonable rates of peritonitis in this study can be explained by a single-center study in which patients were monitored regularly.Also, it can be associated with improving patient education, according to the guidelines.
Our study had some limitations.First, this study had a retrospective design, and data collection was not part of a formal prospective study in which all designable parameters were collected.Second, single-center study results may not apply to the other centers.However, the data are fair and can be checked easily in a single-center study.
The mortality rate of the patient population at our center was similar to Europe and the United States.Cardiovascular diseases and diabetes are the leading causes of death in Turkish PD patients, as in other populations.

Declaration of Interests:
The authors have no conflict of interest to declare.

Funding:
The authors declared that this study has received no financial support.

Figure 1 .
Figure 1.Flowchart of the patients in the study.

Figure 2 .
Figure 2. Effect of initial kidney replacement method on patients' survival.

Table 2 .
Study Parameters of Patients According to Mortality P-values were obtained by comparing living and deceased patients, and values below .05are shown in bold.Variables are demonstrated as median (interquartile range 25-75) or number (percentages).CCr, creatinine clearance; PD, peritoneal dialysis.

Table 4 .
Factors Associated with Mortality According to the Cox Regression Analysis.

Table 5 .
Comparison of Patient Survival Outcomes Under Peritoneal Dialysis
Ethics Committee Approval:Informed Consent: Informed consent was obtained from the patients who agreed to take part in the study.Peer-review: Externally peer-reviewed.Author Contributions: