MTBE: the headache of cleaner air.

Background: Hemodialysis (HD) and peritoneal dialysis (PD) are important renal replacement treatment in end stage renal disease (ESRD), but the comparison of quality of life (QOL) and causes of hospitalisation between the two modalities in China is lacking. In the present study, we compared the two modalities in a multi-center study.


Background
Both hemodialysis (HD) and peritoneal dialysis (PD) are established renal replacement therapies to treat patients with end-stage renal disease (ESRD) worldwide. According to the data of Registry of Dialysis and Transplantation in 2005 in China, there were about 59,000 dialysis patients in this country and 90% of them were on HD. Although HD and PD are thought to provide similar benefits to ESRD patients, the actual large scale comparisons of quality of life (QOL) and causes of hospitalisation between the two modalities is currently lacking in China. In literatures, health related 36 items short form questionnaires (SF-36) were used increasingly to assess the QOL in the studies of ESRD patients [1][2][3][4]. So we employed SF-36 as a tool and investigated the QOL and causes of hospitalisation in HD and PD patients in China based on multicenter data in the present study.

Data collection
Demographic and clinical data of 661 HD patients and 412 PD patients were derived from 10 big hospitals located in different cities in China from Sept, 2004 to Jan, 2005. Demographic information and clinical history were collected through review of medical records and selfreported SF-36 questionnaires. We categorized sociodemographic variables as follows: age, currently marriage status (yes/no), employment status (yes/no). Comorbidity was assessed by using Charlson Cormobidity Index, with higher score reflecting increased severity of disease [5][6][7][8].

Hospitalisation
Retrospective hospitalisation data were collected for year 2004. Among all the patients, 757 patients had complete data and were enrolled for the hospitalisation analysis, but only 309 patients of them were defined hospitalisation, the other were defined non-hospitalisation for access-related hospitalisation (HD patients artery-venous fistula operation, PD catheter implantation) and transplanted-related hospitalisation were excluded. Hospital admissions for a variety of disorders were analyzed [12].

Statistical analysis
Demographic and clinical data were described as mean ± SD for continuous variables. The differences in demographic data and SF-36 scores between PD and HD groups were analyzed by independent t tests for normally distributed variables or nonparametric tests for non-normally distributed variables. The effects of age, gender, diabetic status and Charlson index were adjusted by covariate analysis. To perform covariate analysis, the logarithmically transformed SF-36 scores were treated as dependent variables, while the modality of dialysis was treated as a fixed factor and the age, gender, diabetic status and Charl- son index as covariates. P < 0.05 was considered as statistically significant. Statistical analyses were performed using SPSS 11.0 version.

Comparisons of quality of life
The average scores for SF-36 domains of BP, GH, RE, SF, VT and MH were higher in PD patients than those in HD patients except PF and RP. This difference was still significant even after adjustment of age, gender, diabetic status and Charlson index by covariate analysis (Table 2). However, there were no significant differences between the diabetic patients and non-diabetic patients in terms of physical health and health dimensions, neither between the female and male patients (Table 3 and Table 4).

Analysis of the causes of hospitalisation in dialysis patients
HD patients had higher hospitalisation as compared to PD patients (47.8%vs 29.7%, P=0.000). In HD patients, the two most common causes of hospitalisation were cardiovascular diseases and infection (especially pulmonary infection), while they were infection (especially infectious peritonitis) and cardiovascular diseases in PD patients (Table 5).

The comparison of SF-36 domains between hospitalised and non-hospitalised dialysis patients
Among the two major domains of the SF-36, the physical health dimension scores were also statistically different between the hospitalised patients and non-hospitalised patients, and the ever hospitalised patients had significant lower SF-36 domain scores (containing PF, BP, GH, RE, SF, VT, MH, see in Table 6).

Discussion
Previous study reported that there was no simple answer to the question of which dialysis modality could be expected to provide better quality of life [9,10]. Several studies suggested advantages for PD in some domains [11,12], and HD in others [3] or little difference between the two modalities [13,15].
The current study performed in China showed that PD patients reported better quality of life in mental health dimensions, GH (Combined mental health dimensions and physical health dimensions) and BP of physical health dimensions than HD patients except PF and RP if non-matched demographic data were not adjusted. This difference remained significant after adjustment of the patients' characteristics. It was interesting that we failed to demonstrate that diabetic status affected QOL, which was contrasted to some studies [2], but consisted with our previous investigations [14]. The exact cause for better QOL in Chinese PD patients is not clearly at present, but we believe that the following reasons should be considered. Firstly, the lower body mass index in Chinese patients may be translated as an adequate dialysis could be achieved with a relatively lower dialysis dose. Indeed, there are reports that the survival in Hong Kong Chinese PD patients is better than that in Caucasian patients in Western world, although the former have relatively lower Kt/V [16]. Secondly, it is reported that the micro-inflam-mation state, a predictor of cardiovascular event in dialysis patients [17], is lower in Asian patients than that in Western patients [18], which may be due to the difference in race and/or dietary habits.
Our current study also showed that the RP score in Chinese dialysis patients was low, consisting with the low employment rate (HD 3.2%, PD 7%) observed in this study. This might be due to the following reasons: Firstly, many Chinese patients do not accept the concept of timely dialysis until they suffer more and severer comorbidity; Secondly, the Chinese dialysis patients usually depend on the care of their family members (their children, spouse, sisters and brothers, parents, etc); Thirdly, Chinese dialysis patients are often not re-employed because of their end stage renal disease.
The current study showed that the cause of hospitalisation differed between PD and HD. In our HD patients, cardiovascular disease was most common cause and infection ranked the second. Moreover, majority of cause of hospitalisation were congestive heart failure in HD patients.
Our results were consisted with some previous reports. Rayner [19] reported that cardiac disease was a common cause of death in chronic hemodialysis patients. A subanalysis of the data on cardiac diseases in the Hemodialysis (HEMO) Study by Cheung [20] found that among the total of 1685 cardiac hospitalisations, angina and acute  For PD patients, although majority of Chinese PD patients have been using twin-bag system of Baxter Ltd since late 1990's, infectious disease especially peritonitis remained to be the most common cause of hospitalisation in our patients indicating further efforts are need to decreased the incidence of peritonitis in this patient population.
Kalantar [22] reported that prospective hospitalisations of hemodialysis patients correlated significantly with the SF-36 total score and its two main dimensions. Our retrospective hospitalised patients had worse quality of life than non-hospitalised patients, decreasing cormorbidities should thus be an effective way to improve the quality of life in dialysis patients.
In conclusion, although the present study is not a randomized controlled study and the selection of dialysis modality may have been biased in many aspects, our study indicated that with the current practice in China, PD patients may enjoyed better quality of life than their HD counterparts. Our results also showed that the most common cause of hospitalisation was cardiovascular disease in HD patients and peritonitis in PD patients.