Xerosis elevates the risk of catheter-related infections in peritoneal dialysis patients

Cohort study design

This research was conducted as a multicenter prospective cohort study using data from the Peritoneal Dialysis Outcomes and Practice Patterns Study (PDOPPS) in Korea. Initiated in 2014, the PDOPPS is an international prospective cohort study aimed at enhancing the knowledge of managing PD patients worldwide. The details of the PDOPPS study have been previously described [12].

The current analysis utilized Korea PDOPPS data (626 PD patients, 20 facilities) from phase 2 (2019–2021) [13]. The PDOPPS received approval from both a central institutional review board (IRB) in the USA and the IRB at Soonchunhyang University Cheonan Hospital in Korea (No. 2019-07-045). Eligible patients provided informed consent to participate in the study until the enrollment target was attained, in accordance with the Declaration of Helsinki of the World Medical Association. Patients lacking self-reported data on skin xerosis or those with follow-up periods shorter than 120 days were excluded to ensure a focus on sufficiently observed cases (Fig. 1). All patients were aged 18 years or older at the time of study enrollment.

Figure 1.

Flow chart of the study.

PD, peritoneal dialysis; PDOPPS, Peritoneal Dialysis Outcomes and Practice Patterns Study.

Skin microbiota analysis

We analyzed the skin microbiota database of dialysis patients (hemodialysis, 40 and PD, 36) at Soonchunhyang University Cheonan Hospital from our previous research [14]. Nineteen overlapping patients were concurrently enrolled in both the PDOPPS and the microbiota study. This study investigated the association between chronic kidney disease (CKD)-associated pruritus and skin microbiota. Skin microbiome samples were collected from three sites: the back (sebaceous area), antecubital fossa (moist area), and shin (dry area). The skin sites were swabbed using the NBgene-SKIN (NBG-S22S) skin sampling kit (Noble Bio). Following collection, the samples were transported to the Probiotics Microbiome Convergence Center of Soonchunhyang University (Asan, Korea) for DNA extraction. The study obtained approval from the IRB at Soonchunhyang University Cheonan Hospital in Korea (No. 2021-12-020).

Variables of cohort study

Study coordinators at each participating site abstracted data from medical records for each enrolled patient. Abstracted data included demographics, comorbidities, dialysis status, and clinical outcomes. The initial values of body mass index (BMI), type of PD therapy (either continuous ambulatory PD or automated PD), and laboratory parameters during the phase 2 period were utilized. Furthermore, the catheter-related infection rate was calculated per patient annually. Upon reviewing the prevalence of catheter-related infections, the causative agents were identified. Enteric bacteria included Escherichia coli, Pseudomonas aeruginosa, Serratia marcenscens, etc. Additional details, such as the type of PD therapy during the infection, instances of relapsed infection, recurrent infection, and repeated infection, along with other relevant information, were listed.

The definitions of relapsed infection, recurrent infection, and repeated infection of the PDOPPS study followed the International Society for Peritoneal Dialysis (ISPD) guidelines [6,11]. Relapsed infection refers to an infection episode that occurred within 4 weeks after completing therapy for a prior episode, involving the same organism or a sterile (culture-negative) episode. This classification includes cases where a specific organism recurs, a culture-negative episode precedes an infection with a specific organism, or an infection with a specific organism is followed by a culture-negative episode. Recurrent infection is defined as an infection episode that occurred within 4 weeks after the completion of treatment for a previous episode but is caused by a different pathogen. Repeat infection describes an infection episode that occurred more than 4 weeks after the completion of treatment for a previous infection, involving the same pathogen.

Evaluation of the degree of xerosis and pruritus

In the original data of the PDOPPS study, PD patients reported their severity of skin xerosis or skin pruritus on a 5-point scale: not at all (1), somewhat (2), moderately (3), very much (4), and extremely (5). These scales were grouped into three categories: none (1), low-grade (2–3), and high-grade (4–5).

In the skin microbiota study, pruritus severity was determined with the self-reported questionnaire: the Worst Itch Numeric Rating Scale (WI-NRS). It indicated the degree on a scale from 0 to 10, where 0 represents no pruritus and 10 is the most severe pruritus [15]. Our previous study additionally utilized two multidimensional questionnaires: the 5D Itch Scale and UP-Dial, to validate the reliability of WI-NRS. Yet, we solely relied on WI-NRS results as a criterion for categorizing pruritus severity as used in our previous studies (low-grade pruritus, 0–3; high-grade pruritus, 4–10) [14]. The relative abundance of Staphylococcus aureus was specifically examined at three skin sites.

Statistical analysis

Statistical analyses were conducted using IBM SPSS version 25.0 (IBM Corp.). Categorical variables such as male sex, hypertension, xerosis, and pruritus were expressed as counts (percentages). Continuous variables such as age, BMI, and duration of dialysis were presented as either mean ± standard deviation or median (interquartile ranges), as appropriate. For group comparisons, the chi-square test or Fisher exact test was employed for categorical variables. One-way analysis of variance and the Kruskal-Wallis test were used in continuous variables, except for the microbial analysis. The Wilcoxon rank-sum test was used to compare the relative abundance of skin microorganisms.

For further examination, Cox regression analysis assessed the relation between variables and time-to-event outcomes, providing hazard ratios (HRs) and 95% confidence intervals (CIs). The starting point was the initiation date for follow-up for each patient during phase 2 of PDOPPS. Endpoints were defined as the censoring date derived from the database; date of death for patients who died, the minimum of 7 days after the date of last dialysis, and end of facility follow-up for other patients. Multivariate models were employed for the analysis. Kaplan-Meier plots were generated to examine the impact of skin dryness and skin pruritus on catheter-related infections and S. aureus-associated catheter-related infections, with comparisons made using log-rank tests. Moreover, logistic regression analysis was conducted to analyze the risk factors of xerosis and pruritus, and the results were presented as odds ratios (OR) and 95% CI. Missing values were excluded from the analysis without imputation (Supplementary Table 1, available online).

Patients with xerosis have a higher frequency of catheter-related infection

The characteristics of the patients are detailed in Table 1 and Supplementary Table 2 (available online). Patients with high-grade xerosis tend to have concurrent high-grade pruritus (68.0%), in contrast to those with low-grade (8.1%) or no (0.8%) xerosis (p < 0.001). The evaluation of laboratory parameters revealed that higher serum creatinine (p < 0.001), total calcium (p < 0.05), phosphorus (p < 0.001), and calcium-phosphorus (Ca-P) product levels (p < 0.001) are notable in patients with high-grade xerosis. Additionally, the proportion of patients who own pets was higher in the high-grade xerosis group (19.4%) compared to the low-grade (15.4%) or no (8.4%) xerosis group (p < 0.05). When comparing the incidence rate per patient-years of catheter-related infection as average ± standard deviation on an individual basis during the observation period, patients with high-grade xerosis had a higher incidence rate (0.13 ± 0.36 infections/patient-year) than low-grade (0.08 ± 0.33 infections/patient-year) or no (0.03 ± 0.15 infections/patient-year) xerosis (p < 0.01).

Table 1.

Baseline characteristics of the subjects according to the degree of self-reported xerosis^a

Xerosis and pruritus are risk factors for catheter-related infection

The association between poor skin conditions and the first event of catheter-related infections was assessed by Kaplan-Meier and Cox regression analyses. The median follow-up duration was 706 days (interquartile range [IQR], 365–855 days) in overall patients, with no significant differences between groups according to their degree of self-reported xerosis (no xerosis: 577 days [IQR, 332–854 days], low-grade xerosis: 832 days [IQR, 365–855 days], and high-grade xerosis: 837 days [IQR, 397–855 days]; p > 0.05). During the follow-up period, a total of 64 catheter-related infection events were observed. The Kaplan-Meier curves showed that the first event of catheter-related infections is more likely to occur in patients with high-grade xerosis (p = 0.006) (Fig. 2A) and high-grade pruritus (p = 0.001) (Fig. 2B), compared to low-grade or no xerosis and pruritus.

Figure 2.

Kaplan-Meier plots for the first event of catheter-related infection according to the degree of xerosis and pruritus.

Kaplan-Meier plots estimated the infection rate of catheter-related infection by the degree (none [1], low-grade [2–3], high-grade [4–5]) of (A) xerosis and (B) pruritus, and the Staphylococcus aureus-associated catheter-related infection by the degree (low-grade or none [1–3], high-grade [4–5]) of (C) xerosis and (D) pruritus. A log-rank test was performed to evaluate the significance.

Table 2 depicts the univariate and multivariate Cox regression analysis for the risk factors of a first catheter-related infection episode. In the multivariate Cox proportional hazards model, patients with high-grade xerosis had a 2.71 times higher risk for the infection compared to those with no xerosis (95% CI, 1.19–6.18; p = 0.02) (Table 2), and patients with high-grade pruritus had 2.57 times higher risk of infection compared to those with no pruritus (95% CI, 1.27–5.22; p = 0.009) (Table 2).

Table 2.

Risk of catheter-related infections according to the degree of self-reported xerosis or pruritus

Xerosis and pruritus are associated with a higher risk of Staphylococcus aureus-associated catheter-related infection

The detailed aspects of catheter-related infection are listed in Supplementary Table 3 (available online). Among the total of 97 cases of catheter-related infection, most of the cases were ESIs only (93%). While looking at the etiologic organisms of the infections, interestingly, gram-positive strains are more cultured in patients with high-grade xerosis (60.5%), compared to patients with low-grade (34.0%) or no xerosis (33.2%). Furthermore, S. aureus-associated catheter-related infections accounted for 34.9% of cases in patients with high-grade xerosis (15 out of 43), which was higher than the cases occurring in patients with low-grade (14.9%, 7 out of 47) or no (16.7%, 1 out of 6) xerosis.

To investigate whether high-grade xerosis increases susceptibility to catheter-related infections caused by S. aureus, the Kaplan-Meier curves and Cox regression analysis were utilized. According to the Kaplan-Meier curves, patients with high-grade xerosis exhibited a significantly higher probability of experiencing a first event of S. aureus-associated catheter-related infection, compared to patients with low-grade or no xerosis (p < 0.001) (Fig. 2C). This tendency was also apparent when comparing the degree of pruritus (p < 0.001) (Fig. 2D). In the multivariate Cox proportional hazards model, patients with high-grade xerosis had 5.66 times higher risk for S. aureus-associated infection (95% CI, 1.97–16.30; p = 0.001) (Table 3) than those with low-grade or no xerosis, while patients with high-grade pruritus had a 5.93 times higher risk (95% CI, 2.18–16.15; p < 0.001) (Table 3) compared to those with low-grade or no pruritus.

Table 3.

Risk of Staphylococcus aureus-associated catheter-related infections according to the degree of self-reported xerosis or pruritus

Staphylococcus aureus is more abundant in the skin of patients with high-grade pruritus

The microbial analysis aimed to find out whether the abundance of S. aureus varies with skin conditions in dialysis patients. The result revealed that in the antecubital fossa, a moist area similar to the abdomen, the relative abundance of S. aureus was significantly higher in the high-pruritus group, compared to the low-pruritus group (p < 0.05) (Fig. 3) [16]. Additionally, the amount of S. aureus in the back and shin, the sebaceous area, and the dry area, respectively, was also higher in patients with high pruritus, though not significantly (p > 0.05) (Fig. 3).

Figure 3.

The relative abundance of Staphylococcus aureus according to the degree of pruritus.

Data are expressed as means ± standard deviation. Self-reported pruritus severity (Worst Itch Numeric Rating Scale) was classified into two groups: low-grade pruritus (0–3, n = 26) and high-grade pruritus (4–10, n = 50). p-value was calculated using the Wilcoxon rank-sum test (^*p < 0.05).

Pet ownership, high serum creatinine levels, and high calcium-phosphorus products may be the risk factors for xerosis

Risk factors for xerosis and pruritus in PD patients were assessed by multiple logistic regression analysis (Table 4). After adjustment, having any pets (OR, 1.63; 95% CI, 1.00–2.65, p = 0.05), serum creatinine levels (OR, 1.06; 95% CI, 1.00–1.13, p = 0.04), and Ca-P product levels (OR, 1.02; 95% CI, 1.00–1.04, p = 0.03) remain significantly associated with xerosis. Risk factors for pruritus included diabetes mellitus (OR, 1.54; 95% CI, 1.03–2.31, p = 0.04) and Ca-P product level (OR, 1.03; 95% CI, 1.01–1.04, p = 0.003).

Table 4.

Multivariate logistic regression analysis for the risk factors for high-grade xerosis and pruritus