DAMPAK STATUS GIZI PADA PASIEN PRITONTIS

DENGAN CAPD
Narayan Prasad, Amit Gupta, Raj K. Sharma, Archna Sinha, and Ramesh
Kumar
Department of Nephrology, Sanjay Gandhi Post Graduate Institute of Medical
Sciences, Lucknow, India
Objective: To determine the impact of nutritional status on peritonitis in patients on
continuous ambulatory peritoneal dialysis (CAPD) in a developing country.
Methods: 56 patients with end-stage renal disease on CAPD were randomly selected for
this study. These patients were assessed for nutritional status and peritonitis episodes.
Nutritional parameters were assessed by anthropometry,diet, body mass index (BMI),
Nutritional Risk Index (NRI), serum albumin level, and Subjective Global Assessment (SGA).
Based on SGA, patients were categorized into
either group 1 (malnutrition, n = 31) or group 2 (normal nutritional status, n = 25).
Peritonitis was considered the primary outcome and was compared between the two groups.
Results: Demographic profiles, Kt/V, creatinine clearance, and mean follow-up of the two
groups were similar. Numbe r of peritonitis episodes was significantly higher in patients with
malnutrition (25/31) compared to patients with normal nutritional status (4/25) (p = 0.001).
Mean peritonitis rate per patient per year was also significantly higher in patients with
malnutrition (0.99 1.07) compared to patients with normal nutritional status (0.18 0.42)
(p = 0.007). On univariate analysis, malnutrition based on SGA (p = 0.009), NRI (p = 0.02),
serum albumin level (p = 0.005), and calorie intake (p = 0.006) was a significant
predictor of peritonitis. On multivariate Cox regression analysis, only SGA (p = 0.001, odds
ratio 0.08, 95% confidence interval 0.02 0.36) was found to be a significant predictor of
peritonitis. On general linear model, the observed power of prediction of peritonitis was 0.96
based on SGA. On KaplanMei
Correspondence to: A. Gupta, Department of Nephrology, Sanjay Gandhi Postgraduate
Institute of Medical Sciences, Rae Bareli Road, Lucknow 226014, India. amitgupt@sgpgi.ac.in
Received 1 August 2005; accepted 3 May 2006.
survival in patients with normal nutrition (42 months) was significantly higher compared to
patients with malnutrition (21 months) based on SGA (log rank p = 0.003).
Conclusion: We conclude that peritonitis rate is high in patients with malnutrition and that
malnutrition indices, especially SGA, can predict the peritonitis rate in CAPD
patients. Perit Dial Int 2007; 27:4247 www.PDIConnect.com

KEY WORDS: Malnutrition; peritonitis; Subjective Global Assessment.

utritional indices are significant predictors of patient outcome on peritoneal

dialysis (PD) (1). Malnutrition is a major cause of morbidity and mortality in endstage renal disease (ESRD) patients on PD. Anthropometric and biochemical
evidence of malnutrition is seen in 18% 56% of patients on PD (2,3). Based on
Subjective Global Assessment (SGA), up to 72% of Indian
patients on PD are malnourished (4). Unfortunately, peritonitis remains a major
cause of technique failure, morbidity, and mortality, despite advances in
connectology and technique used for PD (5,6). Although the incidence of peritonitis
has been markedly reduced because

external contamination during the exchange procedure is less likely with disconnect
systems (7), the impact of endogenous factors, including malnutrition, remains to
be determined. The impact of malnutrition on patient morbidity and mortality has
been studied in great detail (1,8), but there is still a paucity of data on the impact of
nutritional indices on peritonitis in these
patients (9). Therefore, we undertook this study to determine the impact of
nutritional status on peritonitis in continuous ambulatory peritoneal dialysis (CAPD)
patients in a developing country.

METHODS
Patients (n = 56) with ESRD were randomly selected from a pool of 196 patients
who started on PD from July 2001 to December 2002 and who gave their consent for
enrollment and follow-up in the study. All these patients had a double-cuffed
straight Tenckhoff catheter inserted by a surgeon under local anesthesia and were
started on UltraBag disconnect system and Dianeal CAPD fluid supplied by Baxter
Inc. (Gurgaon, India) All patients had undergone similar training by the nurses
dedicated to PD at our institute and were started on CAPD after a break-in period of
2 weeks. All patients were prospectively followed from the day PD was started.
Duration of
PD was defined as the period from the day PD was started until the end of follow-up
in this study. These patients were assessed for nutritional status at 1 month after
initiation of dialysis. Nutritional parameters were assessed by anthropometry,
dietary diary of 72 hours, body mass
index (BMI), Nutritional Risk Index (NRI), serum albumin level, and SGA. Body mass
index was calculated as the weight divided by the square of height, and NRI was
calculated by the formula NRI = 1.519 serum albumin (g/L) + 41.7(current
weight/usual weight) (10). Subjective Global Assessment is a valid estimate of
nutritional status for patients treated with PD (1,11). SGA was assessed as per the
dialysis malnutrition score system proposed by KalantarZadeh et al. (12). All seven
features of SGA [weight change, dietary intake, gastrointestinal symptoms,
functional capacity, comorbidity, loss of subcutaneous fat (triceps, biceps,
subscapular), and signs of muscle wasting (temples, clavicle scapula, quadriceps,
knee, interosseous)] were evaluated. Skin fold thickness was measured by fat
caliper in at least three places (biceps, triceps, and subscapular). A score of 1
(normal) to 5 (very severe) was assigned to all seven features of SGA. Total score
ranges from 7 (normal) to 35 (severely malnourished). A SGA score up to 7 was
considered normal nutrition, score 8 14 mild malnutrition, 15 21 moderate
malnutrition,
and 22 35 severe malnutrition. Based on SGA, patients were categorized into
group 1 (patients with malnutrition: SGA score 8 35, n = 31) or group 2 (patients
with normal nutritional status: SGA score 7, n = 25). Kt/V and weekly creatinine
clearance were calculated by PD Adequest software (Baxter). The presence of two of
the following criteria in any combination satisfied the criteria for diagnosis of
peritonitis (13): (1) symptoms of peritoneal inflammation; (2) cloudy fluid with total
leukocyte more than 0.1 109 cells/L, with polymorph response of >50%; (3)
presence of organisms on Gram stain or subsequent culture of PD fluid. Peritonitis
episodes were prospectively obtained from the patients records. Peritonitis was

considered the primary outcome and was compared between the two groups.
Peritonitis rate per patient per year was calculated as peritonitis episodes divided
by patient-years on PD and compared between the two groups.
STATISTICAL ANALYSIS

Data were analyzed by SPSS (10.0) software (SPSS Inc., Chicago, Illinois, USA) and
chi square test was used for comparison between two groups. Students t-test was
used to compare the means between two groups. Univariate logistic regression
analysis was used to see the impact of nutritional parameters on peritonitis.
Multivariate Cox regression analysis was again used to define the cumulative hazard
and 95% confidence interval (CI) and the impact of these nutritional indices on
peritonitis. A general linear model was made to see the observed
power of prediction of peritonitis. KaplanMeier survival analysis was used to
compare peritonitis-free survival between the two groups. Log rank test was used to
determine the significance of survival analysis. A life table was used to compare
survival between different
groups based on malnutrition by SGA. A p value less than 0.05 was considered
significant.

RESULTS
Demographic profiles, Kt/V, creatinine clearance, and mean follow-up of the two
groups were similar (Table 1). Nutritional parameters for the two groups are shown
in Table 2. The proportion of patients developing peritonitis was significantly less in
patients with normal nutritional
status (4/25) than in patients with malnutrition (25/31) by SGA (p = 0.001) (Table 3).
The relative risk of developing peritonitis with malnutrition by SGA was 5.6 (95% CI
2.2 14.3). Mean peritonitis rate was also significantly higher in patients with
malnutrition (0.99 1.07/patient/year) compared to patients with normal nutritional
status by SGA (0.18 0.42 /patient/year, p =0.001). No patients from either group
developed exit-site infection. There was a significant correlation between serum
albumin and peritonitis episodes. Of the 56 patients, 2/12 developed peritonitis with
serum albumin >3.5 g/dL, 19/33 developed peritonitis with serum albumin 2.8
3.49 g/dL, 7/10 patients developed peritonitis with serum albumin 2.1 2.79 g/dL,
and 1/1 developed
peritonitis with serum albumin <2.1 g/dL (p = 0.03).
TABLE 1
TABLE 2
TABLE 3
On univariate analysis, malnutrition based on SGA (p = 0.001), NRI (p = 0.02),
serum albumin level (p =0.005), and calorie intake (p = 0.006) were significant
predictors of peritonitis. However, age, gender, percentage of ideal body weight,
hemoglobin level, diabetes, and duration of CAPD were not found to predict
peritonitis (Table 4). On multivariate analysis, SGA (p = 0.006) was the only
significant predictor of peritonitis. However, NRI (p = 0.61), serum albumin (p =
0.91), and calorie intake (p = 0.98) were not significant factors for predicting
peritonitis.On multivariate Cox regression analysis, SGA (p =0.001, odds ratio 0.08,
95% CI 0.02 0.36) was the only significant risk factor for predicting peritonitis

(Table 5).On analyzing the data using the general linear model, with peritonitis as
source and nutritional indices as dependent variables, the observed power of
prediction of peritonitis was 0.96 based on SGA, 0.81 based on serum albumin, and
0.65 based on NRI. On KaplanMeier survival analysis, median peritonitis-free
survival with normal nutrition (42 months) was significantly higher compared to
patients with malnutrition (21 months) based on SGA (log rank p = 0.003;
TABLE 4
TABLE 5
We also observed that median peritonitis-free survival decreased with increasing
SGA score. Median peritonitis-free survival on the life table was highest with normal
nutritional status (SGA score 7), at 42.3 months; followed by mild malnutrition
(score 8 14), at 27 months, and moderate malnutrition (score 15 21), at 17.4
months; and lowest with severe malnutrition, at
15.3 months (Figure 2).
DISCUSSION
In developing countries, where maintenance hemodialysis centers are scarce and
not sufficient to accommodate all ESRD patients, CAPD has been proven to be a
good alternative to hemodialysis. There are very few state-funded medical
treatment and medical insurance facilities for these patients in India (14). In our
country, the majority of the patients are already malnourished at the initiation of
CAPD (4). The main reason for malnutrition is late referral to nephrologists, with late
start of dialysis therapy. Most hemodialysis units are situated in cities and are not
accessible to patients living in rural areas. In the early 1980s, the combined
financial burden and the considerable patient morbidity and mortality associated
with peritonitis limited the appeal of CAPD as a dialysis modality in ESRD patients
(15). However, technical advances in connecting system, flush-beforefill technique,
and use of UltraBag have decreased the peritonitis rate recently, making CAPD a
feasible alternative to hemodialysis in India. However, peritonitis remains a major
cause of technique failure, morbidity, and
mortality in these patients. Malnutrition has also been reported to be a major
problem in patients starting on CAPD in other developing countries (16,17). Were
and McLigeyo (16) reported
that malnutrition and infection are the main factors limiting the success of renal
replacement therapy in Kenya. Naicker (17) found 76.2% of their CAPD patients
were malnourished, with loss of appetite being an important etiological factor in
South Africa. On the other hand,
the prevalence of mild/moderate or severe malnutrition is lower in developed
countries, as reported from Japan, Europe, and North America (1,2,18). In the
CANUSA study, the prevalence of mild/moderate malnutrition was 44.6% and severe
malnutrition 4.2% (1). Young et al. (2) reported prevalence of mild/moderate
malnutrition at 32.6% and severe malnutrition at 8% in patients from Europe and
USA. Kumano and Kawaguchi (18) reported low prevalences of mild/moderate
malnutrition of 26.2% and severe malnutrition of 3% in their patients. Literature
also reveals the extremely low peritonitis rate of 1 episode in 53.3 patientmonths
in Japan (19). However, peritonitis depends on multiple factors. The lower
prevalence of malnutrition may be one of those factors responsible for lower
peritonitis rates in these countries. In this study, we observed that the risk of
developing peritonitis is higher in patients with malnutrition. The proportion of
patients developing peritonitis was significantly

higher in patients with malnutrition compared to patients with normal nutritional

status. The mean peritonitis rate per patient per year was also higher in patients
with malnutrition compared to patients with normal nutritional status. In the
CANUSA study, after serum albumin, SGA was
the next best nutritional parameter that predicted clinical outcome (1). The NKFDOQI guidelines suggest that SGA can be used to assess nutritional status in PD
patients (20). SGA assessment is simple, very easy, and does not require any
expensive equipment; furthermore, there is no need to solve complicated equations
of creatinine kinetics. It can be assessed by any clinician, dietician, and trained staff
nurse and is therefore very suitable for a country such as India. More importantly,
SGA is a valid estimate of patients nutritional status (1,20). In the
present study, we did not use creatinine kinetics, dualenergy x-ray absorptiometry,
or bioelectrical impedance methods to calculate nutritional parameters because
they
are not feasible in every center, particularly in India and other developing countries.
In the present study, we used NRI for the first time as a nutritional parameter in PD
patients. Nutritional Risk Index appears to be an objective and useful index for
estimating nutritional status because the formula to calculate NRI takes into
account serum albumin, which is an established marker of nutritional status in
dialysis patients (10). We observed that
NRI was the significant predictor of peritonitis on univariate logistic regression
analysis (Table 4). Elias et al. (21) have shown that, after correction of malnutrition,
the protein catabolic rate decreased significantly and that there was a concomitant
decrease in the incidence of new episodes of peritonitis. It is well established that
uremia chronically immunosuppresses
both cellular and humoral immunity. Further, low pH, lactate, and high osmolality of
dialysate fluid in PD impair the phagocytic activity of peritoneal macrophages and
the local defense mechanism (22). Malnutrition, in addition, impairs cell-mediated
(23,24)and humoral immunity (25,26). The absolute total lymphocyte count the
response to mitogens known to activate T cells specif ically has frequently been
demonstrated to be depressed in malnourished patients (23). Due to impaired cellmediated immunity, peritonitis episodes are more frequent in malnourished
patients(24). The serial change in serum albumin level is the reflection of the
nutritional status of these patients. In the present study, we also observed that the
patients with
low serum albumin at 1 month of PD were at risk of peritonitis. Literature reveals
hypoalbuminemia may be a risk factor for peritonitis (27). We also observed a high
incidence
of peritonitis in patients with low serum albumin. Although Marcus et al. (28) have
not shown association between peritonitis and serum albumin, Young et al. (29)
observed that the incidence of peritonitis and length of stay in hospital were greater
in patients who were
hypoalbuminemic, probably due to impairment of the humoral mechanism. It has
also been reported that recurrent peritonitis can lead to malnutrition. Persistent
CAPD-related peritonitis
is associated with severe malnutrition and considerable morbidity and mortality, in
contrast to routine peritonitis (30). In the present study, nutritional status was
assessed at 1 month of initiation of PD. The peritonitis rate per patient year was
significantly higher in malnourished patients; also, the proportion of patients

developing peritonitis was significantly high in patients with malnutrition. Recurrent

peritonitis (more than 2 episodes) was seen in only 2 of 31
patients with malnutrition and in none of the patients with normal nutritional status.
Therefore, recurrent peritonitis does not seem to be a cause of malnutrition in our
patients. We conclude that peritonitis rate is very high in patients with malnutrition
and that nutritional indices, especially SGA, estimated at 1 month of CAPD can
predict peritonitis in ESRD patients on CAPD. We recommend that SGA should be
assessed in every patient who starts on
PD and measures should be taken to prevent malnutrition and peritonitis.