| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
PD IN THE DEVELOPING WORLD |
Pro-Renal Kidney Foundation1 ; Center for Health and Biological Sciences,2 Pontifícia Universidade Católica do Paraná; Nephrology Division,3 Hospital Universitário Evangélico de Curitiba, Curitiba, Brazil
Correspondence to: R. Pecoits–Filho, Center for Health and Biological Sciences, Pontificia Universidade Católica do Paraná, Rua Imaculada Conceicao, 1155, Curitiba, PR 80215901 Brazil. r.pecoits{at}pucpr.br
 |
ABSTRACT |
|---|
 TOP ABSTRACT STUDY POPULATION AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
Objectives: To evaluate patient and technique survival
and to provide an analysis of peritoneal dialysis (PD)-related peritonitis in
25 years of experience in a single center. Study Design: Retrospective study of incident patients
on PD from July 1980 to July 2005. Setting: Single, university based, Brazilian dialysis
program. Patients: 680 patients were analyzed in our study from
July 1980 to July 2005, with a cumulative experience of 15 303 patient-months.
All patients over 15 years of age entering the dialysis program were included
in the study. Patients with less than 30 days of follow-up were excluded.
Biochemical and demographic variables, peritonitis episodes, and patient and
technique survival were analyzed. Results: Mean age at start of PD was 53 ± 16
years; diabetic nephropathy was the main cause of chronic kidney disease.
Cardiovascular disease was the main cause of death (44%); peritonitis was
responsible for 16% of fatal events. The predictors of death in our study were
diabetes [relative risk (RR) 1.23, p < 0.01], advanced age (RR
1.58, p < 0.001), low serum albumin level (RR 1.25, p
< 0.01), and low serum phosphate level (RR 1.39, p < 0.001)
upon starting PD. There were 1048 cases of peritonitis over the 25-year
period, with a significant reduction in incidence after the introduction of
the double-bag system. The number of incident PD patients originating from
hemodialysis increased threefold over the observation period (p <
0.001), with a similar increase in comorbidities over time. Conclusion: In the largest single-center report of PD
experience in Latin America, we describe the overall rate and trends over time
of peritonitis as well as patient and technique survival, which are similar to
previous reports. Significant changes in peritonitis rates and causative
organisms as well as a significant time-dependent increase in high-risk
patients starting PD were observed.
KEY WORDS: Brazil; Latin America; patient outcome.
Brazil is the largest country in Latin America and the fifth largest in the world, with an area of 8547403 km2. According to data from the Brazilian Society of Nephrology, the annual incidence of patients starting dialysis was 181 patients per million population, prevalence in the country was 391 per million population in 2007, and total mortality of dialysis patients was 14.3%. The trends over time of prevalent patients on dialysis over the past 13 years are shown in Figure 1. Penetration of peritoneal dialysis (PD) in Brazil was estimated at 9.2% in 2007; penetration rates over the past 8 years are shown in Figure 2. The Brazilian Society of Nephrology has made an effort to improve the quality of the Renal Replacement Registry; however, dialysis clinics have no obligation to answer questionnaires and, furthermore, compliance has been a major issue. Currently, there are more than 73 000 patients on renal replacement therapy in Brazil (of which 27% are estimated to be diabetics) being treated in 621 dialysis centers (Figure 3), and 7000 of these patients are on PD, representing 9.3% of all dialysis patients treated in the country (1).
|
|
|
Dialysis treatment in Brazil is universally available and primarily (89%) covered by the public health system. Private medical insurance is utilized by 10% of patients on dialysis and full private patients are almost non-existent. Reimbursement for PD is approximately US$1125 and US$1460 per patient per month, respectively, for continuous ambulatory and automated PD, covering PD solutions, PD equipment, and medical and nurse consultations. In hemodialysis (HD), US$960 is spent per patient per month to cover HD maintenance equipment and dialyzers, and medical and nurses' visits. Hospitalization, medication, vascular access, PD catheter insertion, and transport to the center are not covered by the monthly income.
The introduction of PD in Latin America occurred in the early 1980s. The region represents almost one fourth of the world PD population, estimated at almost 150 000 (2), due in particular to the large number of patients treated in both Mexico and Brazil. Curitiba is located in the southern region of Brazil and is currently the seventh largest city in the country, with a population of 1800000 (Figure 3). The history of our center began in the late 1970s when nephrologist Miguel Riella, after training with Drs. Belding Scribner, Fred Boen, and Dimitrios Oreopoulos from Toronto, among other PD pioneers, implemented the first Brazilian PD protocol in 1979. The first patient began PD renal replacement therapy in early 1980 and the center continues to function. In 1984, the ProRenal Kidney Foundation, a philanthropic organization that provides assistance to patients with chronic kidney disease, began its activities and today represents a key source of support for PD patient assistance and research. The foundation supports the work of dietitians, psychologists, dentists, and podiatrists and the PD clinic is located on site. For almost three decades, the center has continued to be a national reference for research and medical assistance, particularly in the area of PD. Preliminary experiences of the center were reported after 1, 2, 5, and 15 years (3–6).
Reports regarding long-term outcome on PD by experienced centers are an important source of reference, allowing for the development of continuous improvement in PD therapy. There are few single-center reports involving large numbers of patients and extended follow-up periods (7–9). Despite the large number of patients on PD in Brazil, data concerning patient outcome in the country are scarce (10). In contrast to other Brazilian centers, our center has maintained a high PD penetration rate over time: generally above 20%.
The aim of the present study was to analyze trends over time and overall trends in patient characteristics, peritonitis, and clinical outcome over a 25-year period in a single Brazilian center.
|
STUDY POPULATION AND METHODS |
|---|
|
TOPABSTRACTSTUDY POPULATION AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
For patient and technique survival analyses, the following parameters were analyzed: age and body mass index at the start of PD, gender, race, underlying renal disease, previous HD, presence or absence of diabetes mellitus, peritoneal equilibrium test, and peritonitis episodes and causative organisms. Selected baseline serum biochemical variables (albumin, calcium, and phosphorus) and hemoglobin levels were also analyzed.
Technique failure was defined as transfer to HD for any reason and peritonitis-related death. Causes of death were divided into five categories: cardiovascular disease, peritonitis related, infections other than peritonitis, other causes of death, and unknown causes.
Peritonitis was defined according to ISPD guidelines by two of the following: (1) abdominal pain; (2) dialysate leukocyte count >100/µL, with at least 50% neutrophils; and (3) positive dialysate microbiological culture. We excluded peritonitis episodes caused by the same organism if it occurred within 1 month after the previous episode. Only glucose-based solutions were utilized. The Y-set and twin-bag system were introduced in 1992 and patients were gradually converted from the standard to the new system. Routine topical use of mupirocin on the exit site was introduced in 1999.
Statistical analyses were performed using JMP version 7.0 (SAS, Cary, NC, USA). Data are shown as mean ± SD, median, or percentages. Data regarding patient survival were censored for patient transfer to HD, recovery of renal function, kidney transplantation, transfer to another center, loss of follow-up, or still on PD at the end of the study. When technique survival was analyzed, transfer to HD and peritonitis-related death were considered final events. Patient and technique survival were calculated using the Kaplan–Meier method. Factor comparisons were performed using log-rank and the chi-square test. With respect to the univariate analysis, all significant variables found were included in a multivariate Cox proportional hazard model to analyze for mortality and technique failure risk factors. Variables recognized to present clinical relevance in the current literature but not presenting statistical significance in our study were also included in the Cox model. A p value less than 0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTSTUDY POPULATION AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
|
To analyze trends in patient characteristics over time, patients were divided into two groups according to the year they started PD: 1980 – 1992 and 1993 – 2005. More patients in the second period group had been previously treated with HD, they were older, and they presented a higher prevalence of diabetes. On the other hand, the prevalence of hypoalbuminemia, hypophosphatemia (most likely related to a better nutritional status), and anemia was lower than that observed in the first period group. Anemia control was better in the second period group (Table 2).
|
PERITONEAL EQUILIBRIUM TEST
Baseline patient data were considered those measurements taken from the
start and to 6 months of PD treatment. Baseline data for dialysate-to-plasma
creatinine ratio at 4 hours were available for 175 patients of the second
calendar period of the study: 16% were high transporters, 44% were
high-average, 35% low-average, and 3% of patients were low transporters.
Diabetic nephropathy was highly prevalent in the high transport group: 65%
versus 37% in high-average, 38% in low-average, and 16% in low transport
patients (p < 0.05). High transporters were also older than the
other three groups, with a median age of 62 years versus 56 (58 in high
average, 52 in low-average, and 57 years in low transport patients; p
< 0.05). Technique survival at 1, 3, 5, and 10 years was 90%, 66%, 39%, and
8%, respectively, in the high transporter group, versus 92%, 82%, 72%, and 52%
in the other groups together (p = 0.05). There were no significant
differences in the patient survival analysis. The 1-, 3-, and 5-year patient
survival rates in high transporters were 85%, 56%, and 34%, respectively,
while in the other three groups together the values were 90%, 59%, and 33%
(p = 0.5).
PERITONITIS
During 25 years, we recorded 1048 episodes of peritonitis in 680 patients,
yielding an overall peritonitis incidence rate of 0.82 peritonitis per
patient-year. Gram-positive organisms were responsible for most (54%)
peritonitis episodes, with Staphylococcus aureus representing 26% and
Streptococcus epidermidis 24%. Gram-negative micro-organisms were the
cause in 13% of episodes and E. coli was the predominant
gram-negative infection at 5%. Fungus-related infection was 5% and negative
cultures reached 26% in our study. Figure
4 shows the calculated peritonitis rates in each of five calendar
periods. Figure 5 reflects the
peritonitis rate by causative micro-organisms over the 25-year period.
|
|
For diabetic patients, the 5- and 10-year survival rates were 15% and 1% respectively, compared to 41% and 15% for nondiabetics (p < 0.001). When age at start of PD was analyzed, the 5- and 10-year survival rates were, respectively, 12% and 1% in patients over 60 years old and 45% and 16% in patients less than 60 years old (p < 0.001). The Kaplan–Meier data for those variables with statistical significance plus the Cox multivariate analysis, where older age was the most important predictive factor of death, showing a relative risk 58% higher in patients over 60 years of age, are shown in Table 3.
|
Univariate analysis of technique failure showed no statistical difference comparing the periods 1980 – 1992 and 1992 – 2005 (chi-square 0.099). In contrast, patient survival was surprisingly better in the initial period (chi-square 3.85); however, on adjusting the data for age, diabetes, hypoalbuminemia, and hypophosphatemia, this difference disappeared (hazard ratio 0.97, 0.81 – 1.17).
|
DISCUSSION |
|---|
|
TOPABSTRACTSTUDY POPULATION AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
Contrasting with previous reports (11), there was a slight predominance of females in our study. Following a trend already reported in other centers of the world (8), mean age and diabetic nephropathy as the underlying disease presented significant increases over the years. Comparing the first calendar period (1980 – 1992) with the second (1993 – 2005), mean age increased 8.12 years (p < 0.01), patients over 60 years represented 40.8% of the second calendar period, and there was a 45.5% increase in the prevalence of diabetic nephropathy (from 25.9% to 37.8%, p < 0.01).
Peritoneal membrane transport characteristics vary widely among patients and can differ significantly in distinct populations (12,13). There are conflicting data as to whether peritoneal transport status per se is an independent risk factor for morbidity and mortality in PD patients (14–16). In our study, we found no evidence of peritoneal transport status as a risk factor for mortality; however, the absolute number of high transporters (only 29 patients) limited the statistical analysis. Even with this limitation, high transport status appears to have a significant impact on technique failure, in concordance with other studies (17).
Peritonitis is still a serious complication and a major cause of mortality and technique failure on PD. Prevalence and etiology vary depending on the region and the time period, and recent guidelines advise defining treatment based on local profile. There was a great reduction in the incidence of peritonitis in our center over this 25-year period. This reduction was associated with advances in connecting device technology, which in our center occurred gradually from 1992 to 1995.
The peritonitis rate decreased over time and the reduced incidence was observed in all types of micro-organisms. However, our peritonitis rates are higher than other Latin America reports, probably reflecting the large number of patients transferred from HD due to vascular access failure and a significant comorbidity profile. In addition, the low socioeconomic and educational levels of our population may have influenced the poor results related to peritonitis. The most striking reduction in the peritonitis rate coincided with the introduction of the twin-bag system (1992 – 1995), which led to a 75% drop in the infection rate (p < 0.001). Gram-positive micro-organisms were responsible for the majority (54.5%) of peritonitis, and this subgroup showed a more intensive decrease than gram-negatives and fungus-related infections over the years, as shown in Figure 5. There was an important change in the causative micro-organism profile. Although without statistical significance, is interesting to note that, with the introduction of topical mupirocin in 1999, we saw an inversion toward a higher proportion of S. epidermidis compared to S. aureus as the main cause of peritonitis episodes.
Patient survival rates in our center are similar to others described in the international literature (7,9). Some well-established predictors of death, such as diabetes mellitus, old age, and hypoalbuminemia, also reached statistical significance in our patients. We also found that a serum phosphate level lower than 3.5 mg/dL was a predictive factor for death, probably reflecting the worse nutritional status in this subgroup of patients (18,19). In our study, other variables, such as hemoglobin and type of solute transport membrane, recognized in the literature as predictors of death were not statistically significant (20,21).
Our technique survival rates were similar to those in studies reported for Latin America but were lower compared to countries with developed economies (7–9, 22–24). Possible explanations for our technique survival can be linked to socioeconomic and health-related issues, both recognized problems affecting developing countries, as recently reported (25–27).
It is noteworthy that there was a large increase in high-risk patients starting PD: there were more diabetics, patients were older at the start of PD, and, in our center, a huge increase in patients with previous HD was observed, from 21% in 1980 – 1985 to 75% in 2000 – 2005. These last data show that a significant number of patients had end-stage renal disease for a long time prior to starting PD, which has an important impact on survival rates. Late referral for patients with chronic kidney disease in our country is one of the principle factors contributing to this situation. The majority of patients in need of immediate dialysis are seen in emergency rooms. Most of these patients will receive a central venous access and will remain on HD. Exposure to months or years of HD in addition to the inherent traditional risk factors will most likely have a negative impact on established known risk factors such as residual renal function and presence of inflammation (28). However, previous HD was not a predictor of death or technique failure in the present study. The main causes of dropout from HD in our center were vascular access failure and severe coronary artery disease.
Unfortunately, as this is a retrospective study, there is an absence of sufficient data to analyze exit-site infections, residual renal function at start of PD, and comorbidities prior to PD. Furthermore, we did not report or quantify dropout and its causes.
In conclusion, the present study represents a large and long-term description of the experience of a single PD center in Brazil. Although patient and technique survival and risk factors are not distinct from previous reports, there has been a profound change in patient profile through the years. This change in patient morbidity may justify the lack of significant improvement in patient and technique survival.
|
DISCLOSURE |
|---|
|
TOPABSTRACTSTUDY POPULATION AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
Received 22 May 2008; accepted 29 October 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTSTUDY POPULATION AND METHODSRESULTSDISCUSSIONDISCLOSUREREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
