1. Introduction
Peritoneal dialysis (PD) is the least used extrarenal cleansing technique; it only involves 11% of dialysis patients, with great variability between the different regions; despite the definite advantages, both in terms of the autonomy it provides and its accessibility to the most fragile patients [1].
Non-infectious complications are secondary to the implementation of the DP catheter in the peritoneal cavity, the increase in intra-abdominal pressure by the dialysate, and the metabolic effects of absorption of glucose and its by-products. Mechanical complications are frequent and represent around 20% of hemodialysis transfers. Prevention, early recognition and proper management of these complications are paramount due to patient-related morbidity and technique failure [2].
In Senegal, PD has been practiced since 2004. It involves around 12% of patients in extrarenal treatment. This low level of patient management by this technique is partly linked to the occurrence of complications. Besides the most frequent and serious infectious complications, non-infectious complications represent a non-negligible cause of failure (8% to 20% transition to haemodialysis) [3] [4]. These complications are mechanical and metabolic.
Since the implementation of this technique in Senegal, no study has focused specifically on non-infectious complications of PD. Therefore, we conducted this study, which aims to determine the prevalence and the types of non-infectious complications in our PD unit.
2. Patients and Methods
This is a retrospective and descriptive study conducted in the peritoneal dialysis unit of the nephrology, dialysis and renal transplant department of the Aristide Le Dantec teaching hospital in Dakar (Senegal), over a period of 5 years, from 1 January 2013 to 31 December 2017. We included all patients who underwent peritoneal dialysis treatment for at least 3 months for chronic kidney disease at stage 5D. PD patients for acute renal failure and patients on vacation in Senegal were not included.
Data were collected from patients’ medical records. For each patient included, the following parameters were studied: age, sex, causal nephropathy, PD procedures and clinical and biological parameters, notably albuminemia, glycemia, total cholesterolaemia, HDL-cholesterolaemia, LDL-cholesterolaemia and triglyceridaemia, were collected.
For each type of mechanical and metabolic non-infectious complication, we collected the average age of patients and the average time to onset compared with the start of PD.
In terms of equipment, the Tenckhoff double-sleeve catheter was used on all patients. The catheter was placed by the surgical method by a non-dedicated urological surgeon. The techniques used were continuous ambulatory peritoneal dialysis (CAPD) and continuous cycling peritoneal dialysis (CCPD). For CAPD, a BAXTER® type disconnectable double pocket system was used. Glucose bags (physioneal®) of 1.36%, 2.27% and 3.86%, icodextrin (Extraneal®) and amino acids (Nutrineal®) were used.
Mechanical complications were defined by the appearance of complications due to the catheter after its insertion and/or due to the presence of intraperitoneal fluid. Metabolic complications were those attributable to PD (hypokalaemia, hyperglycaemia, hypoalbuminemia, dyslipidaemia).
The collected data were analysed by using IBM SPSS Statistics version 25 software. Qualitative variables are presented as a percentage, and quantitative variables are presented as an average plus or minus the standard deviation or with a minimum and a maximum.
3. Results
• Study population
During the study period, 84 patients benefited from the placement of a peritoneal dialysis catheter. Nine patients were not included due to unusable medical records. The study thus involved 75 patients, of whom 57 were on CAPD and 18 were on CCPD. There were 30 men (40%) and 45 women (60%), for a sex ratio of 0.66. The mean patient age was 48.6 ± 14.4 years.
Seventy-one patients (94.7%) were hypertensive, and 7 patients (9.3%) were diabetic. The first causative agent of nephropathy was nephroangiosclerosis in 31 patients (41.3%) (Table 1).
Sixty-six patients (88%) presented at least one non-infectious complication, including 23 men (34.8%) and 43 women (65.2%). The mean age was 50.1 ± 13.9 years. The mean duration of PD was 18 ± 11 months, with extremes of 3 and 56 months.
• Mechanical complications
The number of patients who presented with at least one mechanical complication was 34, i.e. a prevalence of 45.3%. The mean age was 47.7 ± 12.3 years. Table 2 represents the prevalence of the various mechanical complications and their average time of onset compared with the beginning of PD.
• Metabolic complications
The number of patients who presented at least one metabolic complication was 57% or 76% of the patients included. The mean age was 50.8 ± 13.8 years. The distribution of patients according to metabolic complications is shown in Table 3.
4. Discussion
Mechanical complications in PD are frequent and remain highly variable depending on the region of the world, experience of services, type of catheter and insertion technique used.
The prevalence of mechanical complications in our study (45.3%) was slightly higher than that of the Moroccan and Pakistani studies, where the prevalence was 37% and 40%, respectively [5] [6].
![]()
Table 1. Distribution of patients by cause of nephropathy.
![]()
Table 2. Mechanical complications and their average time of onset.
![]()
Table 3. Metabolic complications and their time of onset.
In our series, catheter migration was found in 25.3% of cases and represented the most frequent mechanical complication (55.9% of complications). These results are corroborated by Miftah et al. [4], as well as Flayou et al. in Morocco [7], who noted a respective prevalence of 24.2% (or 65.2% of complications) and 23.6% (or 36.8% of complications).
Blockage of the catheter was noted in 10.7% of our patients. Miftah et al. [5] reported a higher prevalence at 13%. Huraib et al. in Saudi Arabia [8] found an even higher prevalence of obstruction (25% of cases).
The prevalence of hernia in our study was 5.3%, with a predominance of umbilical localization in 75% of cases. Miftah et al. [5], Flayou et al. [7], Villemain et al. in France [9] and Huraib et al. [8] found a higher prevalence at 21.7%, 12.2%, 11.8% and 10%, respectively. In PD this complication generally constitutes the recurrence of a pre-existing hernia and the number increases with age [9] [10]. In our series, patients who presented with hernia had a relatively low average age, lower than that of the other series. Furthermore, none had a history of hernia, hence this lower frequency.
Accidental cutting was found in 8% of our patients. This mechanical complication had the longest average time to onset (13.5 ± 8.9 months).
Miftah et al. [5] found a slightly lower prevalence at 6.5% after a long average delay (25.6 months). Flayou et al. [7] and Raina et al. in India [11] found a much lower prevalence of 3.4% and 2.47%, respectively. This difference in prevalence could be explained by the fact that some of the teams put the catheter in pockets that the patient carried regularly, which made it possible to protect the catheter. The other complicating factors related to the age of the catheter, the use of sharp clamps, the repeated use of disinfectants and the use of radio opaque barium sulphate catheters.
One patient presented with an intestinal perforation. Asif et al. in the USA [12] and Peppelenbosh et al. in Holland [13] found a prevalence of 0.8% and 0.9%, respectively.
Intestinal perforation is an infrequent but serious complication associated with significant morbidity and mortality [14]. One of the risk factors is the presence of intestinal distension, hence the value of preoperative laxatives.
Postoperative wall bleeding and hemoperitoneum were infrequent in our series. These are relatively mild complications that can be avoided by minimally invasive techniques for implanting the dialysis catheter [15] [16]. Indeed, over a period of 10 years, Vaux et al. [17] did not report any bleeding episodes after insertion of the catheter by the fluoroscopic method.
Regarding metabolic complications, the prevalence of hypokalaemia was 55.9%. Chuang et al. in Taiwan [18] noted a similar prevalence (56.8%). Hypokalaemia is a fairly common complication in PD; its prevalence varies between 10% and 58% [19] [20] and is probably influenced by differences in dietary potassium intake between countries.
In our study, the average albumin level was 27.3 ± 4.9 g/l, compared with 32.3 ± 4.3 g/l in a series by Kaysen et al. in the USA [21]. These differences can be explained by the overall nutritional status of patients with end-stage CKD in our context. Patients are diagnosed at advanced stages of the disease, often with poor nutritional status. This already precarious state is aggravated when the patient begins peritoneal dialysis, due to peritoneal albumin losses.
The prevalence of hypercholesterolemia was 59.3% and that of hypertriglyceridemia, 14.8%. The aetiologies of hyperlipidaemia are multifactorial. On the one hand, the significant absorption of glucose from the dialysate constitutes a substrate for the synthesis of lipoprotein, having as a corollary the rise in the plasma lipid level. On the other hand, the increased peritoneal loss of proteins (including albumin and other lipoprotein regulating substances) is the cause.
Limitations of our study lie in its retrospective aspect and the usual bias in connection with this type of study.
5. Conclusion
In our study, non-infectious complications related to peritoneal dialysis were frequent and varied. They were often the cause of definitive transfer to haemodialysis. The establishment of a therapeutic and dietetic education unit is necessary to reduce the prevalence of these complications.
Acknowledgements
Our thanks to the entire urology team and the nurses in our PD unit for their involvement in the management of non-infectious PD complications.