For this retrospective observational cohort study, we analyzed 109 peritonitis cases occurring in 62 PD patients to examine factors related to catheter removal. Our results demonstrated that concomitant tunnel infection and ST time >24 hr were associated significantly with PD catheter removal because of peritonitis. In addition, the PD duration at peritonitis and the first peritonitis episode were detected as independent factors related to ST time longer than 24 hr. Few studies have specifically examined the relation between the actual required time from awareness of cloudy effluent to initiation of treatment and catheter loss.
As described in the ISPD guidelines, successful treatment of PD related peritonitis apparently requires the start of antibiotics administration without delay. The PROMPT study in Australia demonstrated that contact to treatment time (CT time), defined as the time between the first health provider contact and introduction of antimicrobial therapy, is independently associated with PD failure (catheter removal or death) at 30 days (OR[95%CI]: 1.068 [1.01–1.13], p= .02).7 It is particularly interesting that the relation between CT time and PD discontinuation was significant when peritonitis patients initially presented to hospital-based facility, but not to an ambulant care facility.7 The authors implied that presentation to hospitals with differing triage priorities or to those unfamiliar with peritonitis might cause a delay in starting antibiotic treatment. Results from our study showed that ST time >=24 hr is independently associated with catheter removal or death. Results suggest that the patients with suspected peritonitis according to cloudy effluent should visit a hospital as quickly as possible and that medical staff should start antibiotic treatment promptly. At medical facilities, establishing an action protocol for peritonitis and support or training for medical staff who are unfamiliar with PD will be necessary to reduce the time until starting treatment. At our hospital, because the management for all peritonitis cases is determined by the nephrologist from the initial treatment and because an action algorithm for peritonitis is established, there might be less delay in initiating antibiotic treatment after the visit. This might be one reason why the CT time was not associated with catheter loss in our study, unlike earlier reports.
Why was a longer ST time associated with the risk of PD catheter removal or death following peritonitis? A possible explanation for this finding might be the process of microbial biofilm formation along the catheter. The biofilm formation mechanism is described as follows.8,9 1) The immune response to the catheter itself develops the formation of a conditioning film, consisting of leukocytes, macrophages, and mesothelial cells. 2) The bacteria will be encased in multiple layers of extracellular matrix such as extracellular polymeric substances produced by themselves. 3) Inflammatory cells collect around this structure. Biofilm prevents antibiotic delivery to the bacteria. In addition, biofilms act as a reservoir of microorganisms.8 Therefore, biofilm formation might engender refractory or relapsing peritonitis. Although the necessary time for biofilm organization on the PD catheter surface remains unclear, animal studies have shown that Staphylococcus aureus rapidly formed mature biofilm on wounds within 24 hr.10 Therefore, it might be necessary to start appropriate antibiotic treatment within 24 hr of the onset of symptoms, at the latest, for PD-associated peritonitis.
Our results demonstrated that, in addition to ST time, concomitant tunnel infection is independently correlated with PD failure attributable to peritonitis. Several reports have described factors associated with undesirable outcomes in PD related peritonitis. Yang et al. reported that the etiologic source of the infection, concomitant tunnel/exit-site infection, and abdominal visceral catastrophes are likely to be associated with PD catheter loss attributable to peritonitis.11 Moreover, Tunnel infection and exit-site infection are known as major predisposing factors for PD-related peritonitis. Therefore, careful physical and ultrasonographic examination of the catheter tunnel should be performed routinely when peritonitis is diagnosed.6 Early, prompt diagnosis and management of PD catheter-related infection before development of peritonitis are important to prolong PD catheter survival. In another report, Ram et al. identified hypotension, loose stool, and paralytic ileus as risk factors of catheter loss.12 These symptoms might reflect systemic or abdominal manifestations of PD-related peritonitis, leading to unstable circulation and malnutrition. Therefore, early start of treatment before developing systemic or abdominal complications is warranted to overcome peritonitis.
In our retrospective cohort, the first peritonitis episode and PD duration were associated with delayed ST time (ST time >=24 hr). Although all patients were educated at PD initiation about actions to be taken at the onset of peritonitis, it might be difficult to practice those actions when peritonitis occurs for the first time after a long period. Patients with long PD duration tend to self-determine based on their own experiences. Opportunities for re-education about initial actions to be taken at the onset of peritonitis should be provided for PD patients. Although re-training programs reportedly are unable to prevent the onset of peritonitis13, continuous education might contribute to reduction of PD catheter loss after peritonitis.
Comparison of the values obtained for variables between groups with ST time <24 hr (early group) and >=24 hr (late group) revealed that patients in the late group were more likely to use oral antibiotics before visiting a hospital. Although pre-prescription of oral antibiotics might be useful to avoid treatment delay in the case of difficulty of a prompt visit, it might rather delay visiting and initiation of intraperitoneal antibiotics. No evidence was found for associations between the use of oral antibiotics in advance and reduction of the catheter removal rate. From these perspectives, rigorous education about the appropriate use of oral antibiotics is apparently fundamentally important for PD patients.
Detection of the causative organism is important for appropriate treatment of PD-related peritonitis. However, the rate found for culture negative peritonitis was as high as 30% in our study. A possible reason for that high rate is that the detection sensitivity of the causative organisms might be low because microbiological examination of PD effluent using blood-culture bottle was not performed until 2013. Another reason is that, as described in Methods, oral antibiotics were pre-prescribed for peritonitis until 2013. Although patients were educated to take oral antibiotics after draining cloudy PD effluent for sampling, not all patients were able to do so. Oral antibiotics administered before obtaining culture specimens made it difficult to identify causative organisms and to choose the optimal treatment. By correcting those shortcomings above, adopting PD effluent culture using blood-culture bottle and withdrawal of oral antibiotics pre-prescription, the rate of culture-negative peritonitis improved to 12% after 2014 in our hospital.
The present study had some limitations. First, we examined only a small number of patients from a single institution. A larger cohort study must be undertaken to establish predictors associated with catheter removal. Second, because this was a retrospective observational study, the severities of peritonitis between removal and non-removal groups might not be equal, although only one patient died, because of peritonitis. Third, we might not have investigated or collected sufficient data of unknown factors affecting catheter removal.
Despite those limitations, our study presents some clinical implications. In conclusion, patients with initiation of treatment more than 24 hr after awareness of abnormal effluent were more likely to require PD catheter removal because of peritonitis. Long PD duration was related to delayed initiation of treatment. Continuous education for patients with long PD duration about prompt visitation at the onset of peritonitis might be necessary to improve outcomes including PD catheter survival. Although further research is needed, our study suggests several hints at better management to overcome PD peritonitis and to achieve longer PD life.