Seasonal variations of enteric peritonitis in Belgium and France : RDPLF data

Abstract

Summary

Little information is available on the seasonal ecology of germs responsible for peritoneal dialysis peritonitis. We performed a retrospective study based on RDPLF data covering the last 20 years and 20411 episodes of peritonitis.

We show that the percentage of enteric peritonitis is highest in summer, lowest in winter and identical in spring and autumn. This higher proportion of organisms of enteric origin in summer has itself tended to increase in recent years.

We postulate that a food contamination by enteric germs associated with an increased bacterial translocation at the level of the digestive tract itself favoured by constipation, as well as changes of the food nature could be responsible for this phenomenon.

These seasonal variations may suggest that probabilistic initial antibiotic therapy should be adapted in cases of suspected peritonitis before the results of bacteriological analysis.

INTRODUCTION

Peritonitis related to peritoneal dialysis has long been considered as the Achilles heel of peritoneal dialysis. Today, they represent 14% and 26% of the causes of transfer to hemodialysis centers in France and Belgium respectively, and 3% and 2% of deaths[1]. Moreover, the identification of the causative organism varies from 10% to 50% from one center to another[2]. The ISPD has published a recommendation on the first choice of antibiotic before knowing the organism[3], recently synthesized by Taghavi and Dratwa[4]. However, these recommendations do not take into account the time of year of peritonitis occurrence. For example, we do not know whether the epidemics of gastroenteritis each year have an influence on the ecology of peritonitis, and therefore whether the choice of probabilistic antibiotic therapy depending on the time of year should vary. There are previous studies of seasonal variations in peritoneal dialysis-related peritonitis, but these are generally small series from a single center with limited numbers of patients over short periods of time. The RDPLF has been recording peritonitis in French-speaking countries for more than 30 years . Our objective is therefore to evaluate, over a long period of time, the seasonal variations of bacterial ecology during peritoneal dialysis-related peritonitis in Belgium and France, with a particular analysis of seasonal variations of organisms known to be of enteric origin.

Patients and methods

We performed an observational cohort study to measure variations in the bacterial ecology of peritonitis in peritoneal dialysis in metropolitan France and Belgium in the French Language Peritoneal Dialysis Registry (RDPLF). The RDPLF database is declared to the Commission Nationale de L’Information et des Libertés under the number: 11950164795. The data were exported to an independent file after total and irreversible anonymization. As the data were retrospective from a registry, the written consent of the patients was not necessary for the study. The design and functioning of the RDPLF have been described elsewhere[1].

The variable of interest in this study is the germ responsible for peritonitis in peritoneal dialysis, as recorded by the nurses and physicians of the participating centers, based on the results of their respective bacteriology laboratory.

We selected from the RDPLF database the centers in Belgium and France, which have similar climatic conditions; the French-speaking regions in the south were excluded because of very different climatic conditions. We studied all reported peritonitis from 2000 to 2019.

The description of treatment is not available in the database, but we postulated that the average treatment duration is 2 weeks. As the ISPD definition of recurrence[3]is the occurrence of peritonitis within 4 weeks of stopping treatment we defined as recurrence any peritonitis with the same germ occurring within 6 weeks of the date of the previous episode.

In order to keep only the germs clearly identified according to their seasonal frequencies, we deleted the episodes of cloudy fluid related to a hemoperitoneum, an eosinophilic reaction, aseptic chylous peritonitis, peritonitis without identified germ, and recurrences. In total we retained 20411 episodes.

We classified the germs as enteric or not following the classification previously described[5]. The season of reporting of the infections was calculated according to their dates and classified as winter, spring, summer and autumn. The percentages of peritonitis related to the same type of germ were calculated for each season by dividing the number of germs of a given type by the total number of germs encountered during each season.

Patients were grouped according to age (<44 years, 45-64 years, 65-84 years, 85 years and older), type of peritonitis (enteric and non-enteric), modified Charlson score (<4, 4-8, >8)[6], diabetic status, and causative germ (E. coli, enterococcus). In each category the seasonality of enteric peritonitis was studied.

Statistics:

Results are presented as percentages of peritonitis by season. The chi2 test was applied to the observed numbers of peritonitis. A p<.05 was considered statistically significant. Tests performed on Medcalc Statistical Software.

Similar results being observed over the last 20 years and the last 10 years, we limited the study in groups of patients to this second period for which we have the Charlson score of the patients.

RESULTS

Over the 10 as well as the last 20 years, the percentage of enteric peritonitis is lowest in winter, identical in autumn and spring and highest in summer: winter 25%, spring 27%, summer 29%, autumn 27% (χ2=24.1; p<0.001). Seasonal differences were as follows: fall-winter ,p=0.016; fall-spring, NS; fall-summer, p=0.009; winter-spring, p=0.043, winter-summer, p<.001; spring-summer, p=0.008.

Figure 1shows the seasonal variations of enteric peritonitis by 2-year periods for the last 20 years, showing that the rate of enteric peritonitis is always the highest in summer except for the period 2004-2005.

Figure 1.Seasonal peritonitis percentages per period of two years from 2000 – 2020.

Study in subgroups of patients over the last 10 years

Effect of age:

Seasonality is found in the youngest (<44years,χ2=10, p<.02) and 65-84 (χ2=12.4 p<.01) but not in the 45-64 group (χ2=5.8 NS) nor in patients >85 (χ2=6.7 NS)

Effect of comorbidities:

Charlson score does not influence the epidemiology of enteric peritonitis, none of the groups are significant (<4, 4-7, >7)

Effect of diabetes

Seasonality is found in non-diabetics (χ2=14.4 p<.001) but not in diabetics (χ2=5.75, NS)

Responsible germ

No seasonality is demonstrated for E Coli and Enterobacter

DISCUSSION

The data in the literature are poor concerning the seasonal variations

.....