PD relative to non-diabetics on HD (HR: 4.99, CI: 2.13?1.71).DiscussionIn our study, we have demonstrated that CAPD as an RRT option among predominantly rural-dwelling ESRD patients is associated with increased all-cause and infection-related risks of death. An increased risk of death from all-causes has been described in mortality outcome studies among dialysis patients in developed and developing countries. Among CAPD patients in Mexico, 1 and 3- year survival was poor at 67 and 48 GS-9620 site respectively while Stack et al reported increasing GDC-0084 web adjusted relative risks for death among PD patients as duration of dialysis increased with the maximum risk demonstrable 24 months after dialysis initiation (HR 1.37, CI:1.25?.51, p<0.001).[13, 14] However, we could not demonstrate an interaction between modality and duration on dialysis but we noticed that CAPD patients who died had a significantly short duration of follow-up than CAPD patients who remained alive (20.2 ?16.3 vs 31.9 ?22.6 months, p<0.001). The risk for all-cause mortality among CAPD patients increased after adjustment for DM, systemic hypertension, and select baseline parameters (HR 1.98, CI: 1.28?.05, p = 0.002). The modifying effect of DM (as a cause of ESRD and comorbidity) onTable 5. Univariate and Multivariate relative risks of all-cause mortality according to Diabetes status and dialysis modality. Unadjusted model Hazard ratio DM- HD DM + HD DM- PD DM+ PD#Adjusted Model # CI Ref Hazard ratio Ref 1.02 1.63 4.99 CI Ref 0.43?.50 1.01?.621 2.13?1.71*Ref 1.03 1.47 3.0.44?.42 0.94?.31 1.54?.67*adjusted for age, albumin, cholesterol and hemoglobin at baselineRef- reference group * P < 0.001;p< 0.doi:10.1371/journal.pone.0156642.tPLOS ONE | DOI:10.1371/journal.pone.0156642 June 14,8 /Baseline Predictors of Mortality in Chronic Dialysis Patients in Limpopo, South Africamortality in PD patients is well known.[15] Vonesh et al reported an age and co-morbidity adjusted average mortality risk ratio of 1.22 among United States (US) Medicare incident dialysis patients with DM.[16] There was a significant increase in the risk of death among our DM patients on PD (adjusted HR: 4.99, CI: 2.13?1.71, p<0.001) [Table 5]. Differences in patient characteristics at baseline may explain our higher mortality rates. Patients in PKDC were at baseline more malnourished than the US Medicare group with lower mean serum albumin levels. Nineteen percent of patients (19.0 ) within the US cohort on PD had serum albumin level <31g/L in comparison to 56.9 of our patients. Serum albumin is a recognised marker of malnutrition in dialysis patients, and in turn, malnutrition a known predictor of poor outcome in PD. Baseline haemoglobin levels were also lower among our patients at dialysis initiation. Over the decades, infection-related deaths have accounted for the majority of non-CV deaths among dialysis patients.[17, 18] The risk of death from infection-related causes was significantly higher among our PD patients with peritonitis being the commonest cause of infection. The relationship between peritonitis and outcomes in PD is well-established. An analysis of the Australian-New Zealand data base demonstrated the notable role peritonitis played in the mortality of PD patients.[19] Infection rates vary among peritoneal dialysis modalities with the lowest risk associated with automated peritoneal dialysis (APD) systems.[20] In our cohort, CAPD was the PD modality utilized and may have accounted for the observed peritonit.PD relative to non-diabetics on HD (HR: 4.99, CI: 2.13?1.71).DiscussionIn our study, we have demonstrated that CAPD as an RRT option among predominantly rural-dwelling ESRD patients is associated with increased all-cause and infection-related risks of death. An increased risk of death from all-causes has been described in mortality outcome studies among dialysis patients in developed and developing countries. Among CAPD patients in Mexico, 1 and 3- year survival was poor at 67 and 48 respectively while Stack et al reported increasing adjusted relative risks for death among PD patients as duration of dialysis increased with the maximum risk demonstrable 24 months after dialysis initiation (HR 1.37, CI:1.25?.51, p<0.001).[13, 14] However, we could not demonstrate an interaction between modality and duration on dialysis but we noticed that CAPD patients who died had a significantly short duration of follow-up than CAPD patients who remained alive (20.2 ?16.3 vs 31.9 ?22.6 months, p<0.001). The risk for all-cause mortality among CAPD patients increased after adjustment for DM, systemic hypertension, and select baseline parameters (HR 1.98, CI: 1.28?.05, p = 0.002). The modifying effect of DM (as a cause of ESRD and comorbidity) onTable 5. Univariate and Multivariate relative risks of all-cause mortality according to Diabetes status and dialysis modality. Unadjusted model Hazard ratio DM- HD DM + HD DM- PD DM+ PD#Adjusted Model # CI Ref Hazard ratio Ref 1.02 1.63 4.99 CI Ref 0.43?.50 1.01?.621 2.13?1.71*Ref 1.03 1.47 3.0.44?.42 0.94?.31 1.54?.67*adjusted for age, albumin, cholesterol and hemoglobin at baselineRef- reference group * P < 0.001;p< 0.doi:10.1371/journal.pone.0156642.tPLOS ONE | DOI:10.1371/journal.pone.0156642 June 14,8 /Baseline Predictors of Mortality in Chronic Dialysis Patients in Limpopo, South Africamortality in PD patients is well known.[15] Vonesh et al reported an age and co-morbidity adjusted average mortality risk ratio of 1.22 among United States (US) Medicare incident dialysis patients with DM.[16] There was a significant increase in the risk of death among our DM patients on PD (adjusted HR: 4.99, CI: 2.13?1.71, p<0.001) [Table 5]. Differences in patient characteristics at baseline may explain our higher mortality rates. Patients in PKDC were at baseline more malnourished than the US Medicare group with lower mean serum albumin levels. Nineteen percent of patients (19.0 ) within the US cohort on PD had serum albumin level <31g/L in comparison to 56.9 of our patients. Serum albumin is a recognised marker of malnutrition in dialysis patients, and in turn, malnutrition a known predictor of poor outcome in PD. Baseline haemoglobin levels were also lower among our patients at dialysis initiation. Over the decades, infection-related deaths have accounted for the majority of non-CV deaths among dialysis patients.[17, 18] The risk of death from infection-related causes was significantly higher among our PD patients with peritonitis being the commonest cause of infection. The relationship between peritonitis and outcomes in PD is well-established. An analysis of the Australian-New Zealand data base demonstrated the notable role peritonitis played in the mortality of PD patients.[19] Infection rates vary among peritoneal dialysis modalities with the lowest risk associated with automated peritoneal dialysis (APD) systems.[20] In our cohort, CAPD was the PD modality utilized and may have accounted for the observed peritonit.
