Variability of aseptic peritonitis rates in the RDPLF
Keywords:peritoneal dialysis, aseptic peritonitis, bacteriological culture, péritonite aseptique, dialyse péritonéale, culture bactériologique
Aim : Peritonitis rate in the French Language Peritoneal Dialysis Registry (RDPLF) is one episode per 32 patient-months. The purpose of the present study is to evaluate the percentage of germ-free, or aseptic, peritonitis in all centers, and also by center.
Methods : a first group representing all patients treated in metropolitan France in 2017 was selected to identify the rate of peritonitis and the percentage of aseptic peritonitis. Some centers have very low rates and the percentages of aseptic peritonitis would have had little value on small numbers, so we selected a second group of patients treated between 2010 and 2017 and retained centers that had at least 20 peritonitis. The rate of aseptic peritonitis in these centers was calculated individually per center. An additional survey was also carried out at the centers to find out their conditions for collecting and culturing effluent peritoneal dialysis fluid.
Results : In 2017, out of 1071 peritonitis, 17.2% had no germ identified. During the period from January 2010 to December 2017, 6068 episodes of peritonitis were reported, including 954 without any germ identified (15.7%). One hundred and one centers had more than 20 episodes of peritonitis during this period. In these centers, the percentage of aseptic peritonitis has varied from less than 6% to more than 50%. The additional survey revealed that the sampling conditions varied considerably from one center to another and sometimes they even were not known to all members of the same team.
Conclusion : the peritonitis rates in France and the percentage of aseptic peritonitis are in line with international recommendations. However, there is considerable inter-center variability which would require defining the best sampling conditions and the application of good bacteriological practices. It also highlights the interest of taking into account the center effect in statistical models used in studies.