Optimizing Peritoneal Dialysis–Associated Peritonitis Prevention in the United States

Abstract

Peritoneal dialysis (PD)–associated peritonitis is the leading cause of permanent transition to hemodialysis among patients receiving PD. Peritonitis is associated with higher mortality risk and added treatment costs and limits more widespread PD utilization. Optimizing the prevention of peritonitis in the United States will first require standardization of peritonitis definitions, key data elements, and outcomes in an effort to facilitate nationwide reporting. Standardized reporting can also help describe the variability in peritonitis rates and outcomes across facilities in the United States in an effort to identify potential peritonitis prevention strategies and engage with stakeholders to develop strategies for their implementation. Here, we will highlight considerations and challenges in developing standardized definitions and implementation of national reporting of peritonitis rates by PD facilities. We will describe existing peritonitis prevention evidence gaps, highlight successful infection-reporting initiatives among patients receiving in-center hemodialysis or PD, and provide an overview of nationwide quality improvement initiatives, both in the United States and elsewhere, that have translated into a reduction in peritonitis incidence. We will discuss opportunities for collaboration and expansion of the Nephrologists Transforming Dialysis Safety (NTDS) initiative to develop knowledge translation pathways that will lead to dissemination of best practices in an effort to reduce peritonitis incidence.

Importance of Peritoneal Dialysis–Associated Peritonitis

Peritoneal dialysis (PD)–associated peritonitis (peritonitis) increases treatment costs, hospitalization events, and death risk particularly within 30 days after an episode (1–3). Multiple or prolonged peritonitis episodes may lead to deleterious peritoneal membrane structural alterations (4). PD catheter removal occurs in 22% of peritonitis episodes (5). Peritonitis is the leading cause of transition to hemodialysis (HD) among patients receiving PD. Concerns regarding peritonitis may limit more widespread PD utilization by patients and providers. The multistakeholder workgroup Standardized Outcomes in Nephrology–Peritoneal Dialysis identified peritonitis as a core outcome for trials in patients performing PD (6).

Although the development of evidence-based preventative strategies have improved peritonitis rates (7,8), variation still exists in peritonitis incidence and outcomes across PD facilities in the United States (Figure 1). Such variation may relate to case-mix differences in patient populations and PD facility practice pattern variation (9). Taken together, reduction in peritonitis rates may be amenable to preventative strategies targeted at the facility level.

Figure 1.

Distribution of facility peritonitis rate (episodes per patient-year, with 95% Poisson confidence intervals) among 103 US facilities, ranked from lowest to highest peritonitis rates. Analysis of US PDOPPS data from 2014 to 2017. Time at risk began at PDOPPS study entry and ended at departure from PDOPPS for any reason (e.g., death, transfer out of study facility, transfer to hemodialysis, or transplant). Peritonitis episodes occurring within 21 days (all) or 22–50 days (same organism) after a previous episode were excluded. n=954 peritonitis episodes. The median facility peritonitis rate is 0.23 episodes per patient-year (50th percentile); the interquartile range is 0.14 episodes per patient-year (25th percentile) to 0.33 episodes per patient-year (75th percentile). PDOPPS, Peritoneal Dialysis Outcomes and Practice Patterns Study.

No comprehensive, standardized reporting of PD-related peritonitis exists in the United States, with limited information provided in the United States Renal Data System Annual Data Report (10). The Centers for Medicare and Medicaid Services (CMS) considers bloodstream infection rates among patients receiving HD in the ESKD Quality Incentive Program, but no analogous metric exists for reporting of PD-related infections (11). There is great need to (1) develop standardized definitions for peritonitis and a protocol for national peritonitis reporting; (2) use this definition to characterize variability in PD peritonitis rates, microbiology, and peritonitis-related outcomes across PD facilities in the United States; (3) identify patient and PD facility characteristics that affect PD-related peritonitis risk; (4) identify best clinical practices, establish benchmarks, and define PD patient training/education program characteristics to reduce PD-related peritonitis; and (5) work with patients, providers, and policy makers to understand how best to implement these practices. Here, we provide key considerations in developing a standardized reporting framework for peritonitis incidence and rates across facilities in the United States, identify evidence gaps in peritonitis prevention strategies, describe successful multicenter dialysis-related infection reduction quality improvement initiatives, and highlight a strategy for translation of best practices.

Considerations in Developing a Standardized Definition for Peritonitis and Related Outcomes

International Society for Peritoneal Dialysis (ISPD) diagnostic criteria for peritonitis require at least two of the following three factors: (1) PD fluid white blood cell count >100 cells/μl with ≥50% neutrophils, (2) abdominal pain or cloudy peritoneal effluent, and (3) a positive culture for organisms in PD effluent (12). A recently conducted systematic review demonstrated that this definition often was not applied across multicenter randomized controlled trials and observational studies where peritonitis was a reported outcome (13). In many cases, only certain elements of the ISPD definition were used, with the cell count criteria alone being the most commonly applied (13).

Given the large intrapatient clustering of peritonitis episodes, the proportion of peritonitis-free patients per year would be useful to report (12). Absolute organism-specific peritonitis rates should also be reported, including antimicrobial susceptibilities. ISPD suggests an overall facility peritonitis rate of <0.50 episodes per patient-year should be targeted (12). However, a recent study from the Peritoneal Dialysis Outcomes and Practice Patterns Study (PDOPPS) found that >75% of US facilities had a peritonitis rate lower than this benchmark (0.26 episodes per patient-year; 95% confidence interval, 0.24 to 0.27), suggesting the possible need for a more aggressive target for the United States (14). In an effort to develop and implement a national standardized peritonitis reporting tool, key considerations across several domains should be critically examined (Table 1).

Table 1.

Key considerations in standardizing peritonitis definition and reporting in the United States

Challenges	Solutions
Definition
High rate of culture-negative episodes	Include episodes that meet PD effluent cell count criteria alone (>100 cells/μl with ≥50% neutrophils) in the absence of abdominal pain, cloudy effluent, or a positive organism from the culture of the PD effluent
	Consider relapsing episodes of peritonitis as culture-negative episode before or after an episode of peritonitis with a known organism but within 4 wk
	Reduce rates of culture-negative peritonitis by optimizing PD effluent culture and handling techniques
Exclusion of “secondary” or “surgical” peritonitis	Include episodes in the context of a bowel or viscus perforation or evidence of peritonitis owing to an enteric organism in association with compromise or inflammation of the gastrointestinal system
Reporting
Exclusion of early peritonitis before first PD treatment at home	Include and report episodes separately that develop from the time of PD access insertion to before the patient receives PD treatment at home
Exclusion of peritonitis that developed during hospital stay	Report episodes that developed over the course of a hospitalization when the individual did not initially present with peritonitis
Clustering of episodes within the same individuals	In addition to reporting an overall peritonitis rate, include the proportion of patients in the facility that remain peritonitis free
High prevalence of small programs (≤15 patients)	Choose an appropriate reporting interval window given the large fluctuations in rates from month to month among small clinics
Need to identify outbreaks, establish root cause analysis	Include an organism-based classification that considers different potential sources entry into the peritoneal cavity and captures key patterns of antimicrobial resistance
	Capture peritonitis episodes that occur with a concomitant exit site and/or tunnel infection of the PD catheter
Establish benchmarks: current guideline-based target of <0.5 episodes per year may be inappropriately high	Target zero peritonitis episodes. Use standardized reporting metrics to develop facility score cards that can be used for quality improvement initiatives
Outcomes
Outcomes related to peritonitis variably defined and reported	Include peritonitis-related events that result in any of the following:
	Hospitalization (as a cause for admission)
	PD catheter removal with or without transfer to an alternate RRT modality
	Death within 60 d of peritonitis presentation

PD, peritoneal dialysis.

Peritonitis Definition

A significant proportion of peritonitis episodes are culture negative. The most recent culture-negative peritonitis rate for the United States was 0.08 episodes per patient-year (22% of episodes) (14). Culture-negative peritonitis may relate to an inoculum of organisms not large enough to elicit positive culture results, unusual/difficult to culture organisms (i.e., mycobacterial species), antecedent antibiotic exposure, or suboptimal PD effluent handling and culture techniques. The high culture-negative peritonitis rate highlights the diagnostic significance of the PD fluid cell count criterion, and unlike prior definitions, it alone may be considered to meet diagnosis requirements for peritonitis in absence of the other two criteria. Relapsing peritonitis (an episode occurring within 4 weeks of completing antibiotic therapy for a prior episode with the same organism) is often counted as one episode. However, guidelines are unclear regarding how to account for a culture-negative peritonitis occurring within a 4-week period of a prior peritonitis episode with a known organism that should likely be considered an infection relapse.

Peritonitis in the context of a bowel or viscus perforation, or imaging evidence of peritonitis owing to an enteric organism in association with compromise or inflammation of the gastrointestinal tract (i.e., cholangitis, diverticulitis, or ischemic colitis), has variably been counted in facility peritonitis rates. Often labeled as “surgical peritonitis,” these episodes are excluded in some reports, but occurred when patients were receiving PD treatment and so should be reported.

Organism Classification and Organism-Specific Reporting

Reporting absolute organism-specific rates is important because organisms such as coagulase-negative Staphylococcus or Staphylococcus aureus may point to touch contamination-related events, poor procedural technique, or PD catheter colonization, all of which may be amenable to preventative interventions to achieve aseptic PD exchange procedures. In contrast, enteric organisms may signal a bowel origin and, occasionally, are indicative of bowel compromise. Although organism-specific characterization of peritonitis has been distinguished into Gram-positive and Gram-negative categories, this is misleading because Gram-positive peritonitis, such as enterococcal species, originates from a bowel source. In contrast, Gram-negative peritonitis owing to pseudomonal species may point to catheter colonization issues. A standard classification for organism-specific reporting of peritonitis is needed. In particular, which organism-specific rates to report (including antimicrobial resistance patterns, e.g., methicillin-resistant S. aureus) should be specified to target preventative efforts to reduce peritonitis related to certain organisms/organism categories and readily identify organism-specific outbreaks.

Time at Risk in Calculating Peritonitis Rates

Time at risk classically begins when a patient initiates PD training (12). Therefore, peritonitis episodes that develop during PD training or immediately after PD catheter insertion (but before the patient is receiving PD treatment at home) may be variably reported and not captured in the existing definition. These episodes should be reported as they may carry different risk factors and may relate to periprocedural practices at the time of PD catheter placement (e.g., use of periprocedural antibiotic prophylaxis during PD catheter insertion) (9,12,15,16). Peritonitis rates should be reported separately for peritonitis episodes that occur before the first PD treatment at home and those occurring once the patient receives PD treatment at home.

It may be advisable to simplify the time at risk calculation. For monthly calculation of catheter-related bloodstream infections under the Centers for Disease Control and Prevention (CDC) dialysis event surveillance system, facilities report the number of HD outpatients who dialyzed in the facility on the first 2 working days of the month to estimate the time at risk for infection (17). Such an approach, however, may over- or underestimate peritonitis rates because of the flux in clinic size over a month, and yield highly variable monthly peritonitis rate estimates, particularly in small PD facilities (27% of patients receiving PD in the United States receive treatment in clinics treating ≤15 patients receiving PD) (18). These issues could be addressed by averaging different point-prevalent census counts across different days in a month and averaging rates over several months to provide more stable peritonitis rate estimates for small PD facilities.

Peritonitis-Related Event Reporting

It is important to determine which peritonitis-related events and outcomes to report as part of a possible national reporting framework, and how to define these events/outcomes. Important events include catheter removal (which may or may not include a permanent transition to HD), hospitalization, and death. Definitions for each of these events have been variably reported. Patients may be hospitalized for peritonitis treatment or develop new onset peritonitis in the hospital, events that are important to distinguish for quality improvement purposes. Differences in the timing and rates of return to PD after catheter removal across PD facilities must be considered in establishing a definition of peritonitis-related PD discontinuation. Capturing peritonitis that occurs in association with a concomitant PD catheter exit site or tunnel infection should be considered in developing targeted prevention efforts, but may be challenged by the lack of robust definitions and their uniform application across facilities.

Challenges in Operationalizing Standardized Reporting of Peritonitis across Facilities in the United States

Operationalizing standardized reporting of peritonitis rates across US facilities is challenged by several factors: (1) a significant percentage of patients receive care in small PD programs (fewer than ten patients), so comparing peritonitis rates across facilities in a valid manner may have limitations and may require either a longer period of observation or a minimal clinic size to do so; (2) variability in nurse/nephrologist training and clinical competence in diagnosing and treating peritonitis may further challenge accurate reporting, coupled with time-constrained nurses entering events in the patient record and adhering to the data collection protocol (these challenges could be overcome with enhanced training and education); and (3) absence of seamless data transfers between hospitals and dialysis facilities make capture of peritonitis events that developed in a hospital or required hospital admission more challenging, and given that 69% of patients with prevalent ESKD in the United States are cared for by large dialysis organizations (19), collaboration and engagement with these organizations will be critical.

Success of Standardized Reporting of Dialysis-Related Infections

Although it has not been directly utilized for PD, a critical aspect of this was use of standardized infection outcome measures (20,21). The National Healthcare Safety Network (NHSN) is a surveillance system developed and maintained by the CDC to track various health care–associated infections—a model for how infection data can be effectively collected from virtually all dialysis centers. A critical aspect of this was use of standardized infection outcome measures (20,21). In 2012, CMS incorporated reporting bloodstream infections in patients receiving HD to the NHSN by all US HD centers as part of the ESKD Quality Incentive Program (22). Many centers reduce data collection burden using electronic medical record data extracts for data submission. In a study that preceded this, the collection and sharing of surveillance data with frontline dialysis care staff prompted improvements in infection rates, in the absence of other interventions; reduced rates have been sustained over time (23–26).

National standardized peritonitis reporting in Australia and New Zealand has led to a reduction in their peritonitis rates. In 2003–2004, the peritonitis rate in Australia was 0.63 episodes per patient-year, below the ISPD target at that time (0.67 episodes per patient-year). Uncoordinated attempts to reduce peritonitis rates had little sustained change. To gain a better understanding of peritonitis within Australia and New Zealand, a peritonitis registry was added into the Australia and New Zealand Dialysis and Transplant ESKD registry to collect detailed data of each peritonitis episode, including treatment and outcomes. This enabled identification of factors associated with higher and lower risk of infection and outcomes. Wide variation was observed in peritonitis rates across PD units because of practice pattern variation, particularly practices divergent from existing guidelines.

The Australian peritonitis registry–facilitated interventions aimed at raising awareness and education regarding best practices for peritonitis prevention and treatment through publications, formal education of trainees, and feedback of peritonitis rates to facilities. In 2011, the first substantial improvement in peritonitis rates was noted (down to 0.4 episodes per patient-year). Further improvements are being targeted with additional explorations of the registry (27,28).

Evidence Gaps in Peritonitis Prevention Strategies

Advances in PD connection systems, including the use of double-bag Y systems and flush-before-fill approaches, have led to significant reductions in peritonitis risk (29–32). Peritonitis preventative strategies with the strongest evidence base, and endorsed by the ISPD, have largely focused on reducing risk of colonization and infection of the PD catheter and exit site, as well as on recommending procedures that facilitate continued performance of aseptic PD exchanges (12) (Table 2).

Table 2.

Potential peritonitis prevention strategies

Strategies	Potential Interventions	Grade of Evidencea
Before peritoneal dialysis initiation
Prophylactic antibiotics at the time of PD catheter insertion	Glycopeptide or first-generation cephalosporin	1A
PD catheter design, configuration, and method of insertion	Double- versus single-cuff downward-directed catheter tunnel	No grade
Optimal patient and care partner training and education, including hand hygiene and aseptic technique	Standardized curriculum	Under evaluation
After peritoneal dialysis initiation
PD solution type/modality	Low-GDP neutral pH PD solutions, glucose-based polymers, CAPD versus APD	No recommendation
PD connectology	Flush before fill	1A
Treatment of hypokalemia	Dietary and/or supplementation	No recommendation
Bowel hygiene	Treatment of constipation/diarrhea	No recommendation
Application of antibiotic cream/ointment to the PD catheter exit site	Mupirocin or aminoglycoside	1A
Exit site care	Cleansing twice weekly and after getting wet (i.e., shower)	1C
Antimicrobial prophylaxis	Antifungal treatment (during antibiotic therapy)	1B
	Colonoscopy/lower endoscopy	2C
	Genitourinary/gynecologic procedures	2D
	Wet contamination	Not graded

PD, peritoneal dialysis; GDP, glucose degradation products; CAPD, continuous ambulatory peritoneal dialysis; APD, automated peritoneal dialysis.

^aAs per International Society of Peritoneal Dialysis guidelines (12).

How patient and care partner training strategies are executed remains poorly understood, but is an important line of investigation to reduce peritonitis risk. Patient education around hand hygiene can dramatically reduce touch contamination. A multicenter survey demonstrated that predialysis training, home visits, and retraining were associated with reduced peritonitis rates, and another study demonstrated lower peritonitis and exit site infection rates with the use of an adult learning theory–based curriculum compared with conventional methods (33,34). There are no studies evaluating the utility of retraining and its frequency and timing on peritonitis risk. Other studies have not found a consistent relationship between the intensity of patient training and peritonitis risk. However, a recent Brazilian cohort study found higher peritonitis risk among patients and care partners who received ≤15 hours of training compared with >15 hours (35–37). An observational cohort study in Hong Kong found that patients who received training by nurses with more than 3 years of experience had a higher risk of developing Gram-positive peritonitis compared with those with <3 years of experience, suggesting the potential importance of maintaining competencies among long-term trainers (38). Although the ISPD has set a formal PD curriculum and PD syllabus for patient training (39), the utility of its use on peritonitis risk has not been formally assessed. In this regard, the Targeted Education Approach to Improve Peritoneal Dialysis Outcomes Trial is a cluster-randomized trial evaluating whether standardized training modules on the basis of ISPD guidelines targeting both PD trainers and patients result in a longer time to the composite end point of exit site infection, tunnel infection, and peritonitis in patients on incident PD compared with existing training practices (40). The notion that different learning styles modify the risk of adverse event rates among patients on home dialysis suggest that optimal patient and care partner training also will require personalized training interventions (41). Such an individualized approach to patient training will need to consider engagement and leadership from providers; human factors engineering; patient and care partner motivation and activation; and customization for different learning styles, education, and cultural backgrounds.

There are limited targeted interventions aimed at reducing Gram-negative peritonitis mediated largely by enteric organisms. This is salient because these organisms are more likely to lead to antibiotic treatment failure necessitating PD catheter removal compared with their Gram-positive counterparts (42). Although there is some suggestion that bowel hygiene, avoidance of hypokalemia, and use of prophylactic antibiotics at the time of lower endoscopic procedures may lower the risk of enteric peritonitis, the evidence base to support these practices is weak and requires further prospective evaluation (43).

Implementation of Best Practices for Peritonitis Prevention

A recent survey across PD facilities enrolled in the PDOPPS found that only 64% of US medical directors reported that antibiotics were given before PD catheter insertion, the lowest proportion of all enrolled countries and for which exists a grade 1A recommendation by ISPD guidelines for reducing peritonitis risk (Table 2). In contrast, 91% of US facilities endorsed using an antibiotic cream or ointment at the exit site, another grade 1A peritonitis prevention recommendation endorsed by ISPD guidelines.

Why some practices are better adhered to versus others is not fully understood and may relate to the variability and availability of systematic treatment protocols, lack of knowledge regarding best practices by providers, and location of care, which for PD catheter insertion, often occurs outside the facility, away from the purview of the PD facility treatment team.

Operating under the hypothesis that more uniform care in pediatric dialysis facilities would lower peritonitis rates overall, the Standardizing Care to Improve Outcomes in Pediatric ESKD (SCOPE) Collaborative was launched in 2011 (44,45). SCOPE, a quality transformation collaborative operated by the Children’s Hospital Association, uses quality improvement methodology to track peritonitis rates across facilities and increase adherence with standardized care bundles focused on PD catheter insertion, patient and caregiver(s) training, and routine follow-up care (44,45). SCOPE centers demonstrated a significant increase in the likelihood of adherence with the follow-up care bundle. The follow-up care bundle requires the dialysis team to review aspects of hand hygiene, exit site care, and aseptic technique with the patient/caregiver at monthly visits, and that the patient/caregiver demonstrate competency with these procedures every 6 months (44,46). The follow-up bundle also requires the PD exit site be scored using an objective scoring tool, that touch contaminations are treated according to ISPD guidelines, and that the patient/caregiver undergo retraining after a peritonitis episode (45–47). As a result, SCOPE demonstrated an associated large decrease in the mean peritonitis rate from 0.63 episodes per patient-year in the 13 months before its launch, to 0.42 episodes per patient-year at 36 months postlaunch (P=0.03) (18). SCOPE currently has 51 participating centers in the United States demonstrating >80% adherence with the follow-up bundle and a continued reduction in the mean peritonitis rate (45). SCOPE has prevented nearly 600 episodes of peritonitis at participating centers, avoiding over 340 peritonitis-related hospitalizations and saving over $10 million in health care costs (45). SCOPE continues work to identify other care practices and modifiable risk factors for peritonitis.

Across adult PD facilities, an opportunity for analogous and additional activity may exist in collaboration with the NTDS initiative. NTDS is an American Society of Nephrology project, funded by the CDC, working since 2016 to target zero infections in US HD facilities (48). NTDS has created a culture of safety with inspired, engaged leadership to change culture and has developed a leadership academy for dialysis facility medical directors and nurse managers, reaching over 750,000 professionals and patients to disseminate knowledge via publications, webinars, seminars, focus groups, and social media. In 2019, NTDS completed an infection prevention curriculum aimed at nephrology fellows in a special edition of NephSAP, a nephrology self-assessment program (49). Adherence to infection prevention best practices is additionally being facilitated by review and, possibly, redesign of dialysis facility processes using human factors engineering. Although the current CDC charge to NTDS is restricted to in-center HD, the methodology and approach can be applied to reduce infections for home-based HD and PD.

Conclusions

With increasing uptake of PD in the United States and initiatives likely to further expand home dialysis utilization, a national reporting system for PD peritonitis will foster an environment of infection prevention, transparency, and accountability (43). Such a reporting system has the capacity to readily identify evolving trends and potential outbreaks in peritonitis, as well as help devise national quality improvement initiatives. This effort will require multistakeholder involvement and support, including patient advocacy groups and government bodies to develop standards for defining and reporting peritonitis, collaboration with dialysis organizations, integration within electronic medical records data capture systems, and development of mechanisms for ongoing data validation. Concurrently, identification of best practices in peritonitis prevention and further evidence generation are needed. Robust knowledge translation pathways similar to those from NTDS and the SCOPE Collaborative are needed at PD clinics across the United States. We hope that the increase in PD utilization is realized in the United States in conjunction with reductions in incidence of peritonitis and peritonitis-related adverse events to make this dialysis modality as safe as possible.

Disclosures

N. Boudville has reported receiving speakers’ honoraria; unrestricted educational grant for research projects; and travel grants for workshops from Baxter Healthcare. He has also received unrestricted research grants and travel grants from Amgen and Roche Pharmaceuticals. A.S. Kliger is funded by the American Society of Nephrology to chair the Project Committee for Nephrologist Transforming Dialysis Safety, and by the National Institute of Diabetes and Digestive and Kidney Diseases to chair the Hemodialysis Novel Therapies. A.M. Neu has served on advisory boards for Relypsa and Reata; has participated in clinical research studies sponsored by Amgen, Roche, Retrophin, and Horizon; and serves as faculty lead for the Children’s Hospital Association’s Standardizing Care to Improve Outcomes in Pediatric ESRD Collaborative and as a committee member for the American Society of Nephrology’s Nephrologist Transforming Dialysis Safety. J. Perl has received speaking honoraria from AstraZeneca, Baxter Healthcare, DaVita Healthcare Partners, Dialysis Clinics Incorporated, Fresenius Medical Care, and Satellite Healthcare, and has served as a consultant for Baxter Healthcare, DaVita Healthcare Partners, Fresenius Medical Care, and LiberDi. M. Schreiber Jr. reports receiving personal fees and other from DaVita Kidney Care Inc. during the conduct of the study and personal fees and other from DaVita Kidney Care Inc., outside the submitted work. B.A. Warady is faculty lead for the Children’s Hospital Association’s Standardizing Care to Improve Outcomes in Pediatric ESRD Collaborative and reports research/registry grants from Baxter Healthcare and the International Society for Peritoneal Dialysis during the conduct of the study. All remaining authors have nothing to disclose.

Funding

Funding support for this article was provided by the Agency for Healthcare Research and Quality, US Department of Health and Human Services (grant 1R01HS025756-01 to Arbor Research Collaborative for Health, co-Principal Investigators: J. Perl and R.L. Pisoni).