Background
Chronic kidney disease (CKD) is an important public health problem which is increasing in terms of incidence and prevalence. The worldwide prevalence is 10–13% [
1,
2] and similar estimates are reported for Africa (13.9%) [
3]. End-stage renal disease (ESRD) is a serious complication of CKD and requires renal replacement therapy (RRT) in the form of haemodialysis (HD), peritoneal dialysis (PD) or transplantation as the mainstay of treatment.
Of the more than 1.8 million patients worldwide on dialysis, less than 5% are in Africa where access to RRT is dependent on very limited government support [
4,
5]. The dialysis rates across Africa are less than 20 per million population (pmp) as compared to a global prevalence of 223 pmp. In Africa, 97% of dialysis patients are on HD [
6]. South Africa has a higher proportion of patients on PD. According to recent registry data 71.8% of patients receiving RRT are on HD and 13.5% on PD [
7].
RRT should not only prolong life but also sustain quality of life [
5,
8]. Transplantation results in improved survival, lower costs and better quality of life [
9‐
14] but transplant rates in Africa are very low, averaging only 4 pmp [
15] and decreasing. In South Africa, the proportion of RRT patients with a functioning transplant has decreased from 55.5% in 1994 to 14.7% in 2014 [
7,
16]. The low transplant rates and increasing numbers of patients requiring RRT result in longer durations on dialysis, emphasising the importance of assessing and optimizing the quality of life of our patients on chronic dialysis.
Health-related quality of life (HRQOL) has become increasingly important as an outcome measure of RRT. The traditional focus on the improvement of survival has recently shifted to include a much stronger emphasis on quality of life [
8]. HRQOL represents the impact of the disease or its treatment on the subjective feelings of patients about their physical, mental, spiritual, emotional, social and functional wellbeing [
17]. HRQOL deteriorates as kidney function worsens [
18] and is an independent risk factor for mortality in dialysis patients [
19]. The Kidney Disease Quality of Life (KDQOL) questionnaire [
20] has been validated for use as a disease-specific measure of quality of life. It combines the SF-36 instrument with kidney disease specific items [
21,
22].
Quality of life among dialysis patients has been shown to be lower as compared to pre-dialysis CKD patients [
23], the general population [
24] and other chronic diseases like congestive heart failure, diabetes, depression and even cancer [
25].
In the large Dialysis Outcome and Practice Pattern study (DOPPS) in the United States [
26] poor scores in the physical component of the HRQOL were associated with increased mortality and increased risk of future hospitalisation.
The quality of life of patients on PD has been reported to be better than for those on HD in some studies [
27‐
30] but others, including a systematic review [
14,
28,
31,
32], have reported no difference between the two treatment modalities. PD patients tended to have lower scores in the role-physical and bodily pain domains whereas HD patients had lower scores in the emotional component [
33]. Low quality of life at the initiation of RRT is associated with increased hospitalisation and higher mortality, emphasizing the need for early interventions [
34,
35].
Very few studies have been conducted comparing the quality of life of HD and PD patients in South Africa [
28,
36,
37]. A study in the Western Cape, South Africa, reported low HRQOL in both dialysis modalities with no difference between HD and PD [
28]. Another South African study [
37] which included dialysis patients from both the private and public healthcare sectors also reported compromised quality of life in both modalities although HD patients were found to have increased vitality and better physical component scores.
There are no qualitative data available from these studies to provide a richer interpretation of the numbers reported. We therefore sought to study the quality of life of patients treated with PD and HD using a comparative mixed methods approach. Focus group interviews have the potential to generate rich data based on the synergy of group interaction and allowed us to ascertain our patients’ perspectives about factors affecting their quality of life.
Methods
We conducted a cross-sectional study involving both qualitative and quantitative methods. Tygerberg Hospital is a public-sector 1384-bed teaching hospital in Cape Town, South Africa. The numbers of patients on RRT is limited to approximately 65 patients on HD and 55 patients on PD due to resource constraints. Only patients who are transplantable are accommodated and the renal replacement programme operates on a PD-first policy as there are limited numbers of haemodialysis slots available. A few patients are accommodated closer to their homes at satellite HD units in the nearby towns of Paarl, Worcester, Hermanus and Vredenburg.
All our HD patients receive three 4-hour, in-centre, treatments per week. High-flux dialysers and bicarbonate-based dialysate solutions are used. Continuous ambulatory peritoneal dialysis (CAPD) using a twin-bag system is the sole form of PD, with the majority of PD patients performing four 2-litre exchanges per day. None of our patients have access to a cycler for automated PD because of the substantially higher costs of this treatment option. Dialysis access is achieved via arterio-venous fistulae or tunnelled catheters for most HD patients. For the PD patients, Tenckoff catheters are inserted at the bedside, or in theatre by the surgical team in cases where there are likely to be intra-abdominal adhesions. All patients on chronic dialysis have access to erythropoietin and intravenous iron as required. Unfortunately, most patients present late and there are usually few opportunities for pre-dialysis counselling.
Dialysis patients were recruited from June 2015 and March 2016 if they had been stable on a treatment modality for more than three months. Participants with other chronic illnesses such as malignancy, chronic liver disease or stroke were excluded from the study. Patients admitted in the previous month for any acute illness were also excluded. Participants completed a validated, pre-tested, self-administered questionnaire, the KDQOL-SF 1.3 [
20]. This instrument includes 43 items which are focused on the health-related concerns of patients with renal disease on dialysis (Table
2, first column) and 36 generic core items (Table
3, first column). Raw scores on the KDQOL-SF 1.3 are transformed to a score ranging from 0 to 100 with higher scores representing better perceived QOL.
Data for quantitative analysis were captured using REDCap [
38] and analysed with Stata/SE 14 [
39]. Data were summarized as means ± standard deviations, or medians and interquartile ranges if not normally distributed. The chi square and Fisher’s exact tests were used for categorical variables. Student’s
t test was used to compare continuous variables, and the Wilcoxon sign rank test when not normally distributed. A
P-value of <0.05 was considered statistically significant.
The qualitative phase of the study involved six hour-long focus group interviews [
40], each conducted by an experienced registered nurse without any connection to the RRT programme. The interviews included three groups comprised of only HD patients and three groups of only PD patients. Each group included approximately six patients, with the participants in each group having been purposively sampled to ensure variation based on sex, race, language and dialysis vintage [
41]. English and Afrikaans
1 were the languages used; the facilitator was fluent in these languages. Prompts were used and leading questions avoided allowing participants the freedom to express themselves.
The interviews continued until data saturation was reached. Interviews were recorded electronically, transcribed verbatim and then analysed using thematic analysis with a view to identifying recurring patterns in the texts [
42]. Accordingly, the transcripts were coded and subsequently thematic labels assigned manually [
41].
Discussion
The use of focus group interviews allowed us to explore patients’ perspectives and provided a rich source of qualitative data to illuminate the scores obtained from the questionnaires. We found no difference in the overall quality of life between HD and PD patients as measured by KDQOL 1.3 scoring. This finding is similar to that reported by other studies, including two conducted in South Africa [
14,
28,
31,
32,
37]. Our overall QOL scores were somewhat lower than those reported from another public sector hospital in Cape Town [
28] and comparable to a South African study which included patients from both private and public healthcare sectors [
37].
Regarding the domains within the KDQOL 1.3 questionnaire, we found differences in the scores for specific domains between patients on the two dialysis modalities. PD patients had lower scores for symptoms, energy/fatigue, sleep and social function but better scores for work status and dialysis staff encouragement. The data which emerged from the focus interviews have shed more light on some of these findings.
The patients on PD verbalised more symptoms related to the ESRD and the complications of their dialysis modality. Peritonitis and catheter exit site infection was a common concern while access infections were not mentioned by the HD patients at all. Pruritus was a notable symptom which affected sleep in the PD patients but was not mentioned by the HD patients. Pruritus is common in dialysis patients and is associated with poor sleep, poor QOL and increased mortality [
44]. Sexual dysfunction, including decreased libido and erectile dysfunction, were reported by patients on both modalities. The presence of the Tenckhoff catheter protruding from the abdomen negatively affected the body image and sexual function of the PD patients. HD patients reported symptoms suggestive of intradialytic hypotension and orthostatic hypotension after dialysis. These symptoms are also predictive of increased mortality [
45]. Longer dialysis vintage may be associated with better QOL [
46]. Our PD patients had a lower vintage and their symptom burden may reflect the initial difficulties in adjusting to life on dialysis, lower haemoglobin and albumin concentrations, as well as the occurrence of complications such as peritonitis and catheter exit site infection.
The lower social function score in the PD group was contrary to our expectation that PD patients would be more independent and therefore more likely to enjoy an active social life. This surprising finding was clarified in the focus group interviews where it became clear that the fear of developing peritonitis caused many patients to sacrifice social interactions to be able do their PD exchanges at home and at the specified times. Some felt that their home was the only safe place to do their exchanges. HD patients felt that their social interactions were limited by the need to remain close to their dialysis centres.
PD patients had better work status scores but nevertheless reported difficulties in finding employment and difficulties in doing bag exchanges while at work. They perceived their doctors and nurses as having unrealistic expectations of their ability to find and maintain employment. The treatment modality was considered to provide flexibility of lifestyle and leave them relatively independent but, in reality, this was mostly not the case. Their high symptom burden and the lack of support by employers in terms of providing time and space to do their bag exchanges made working very difficult. In contrast, the HD patients were generally understood to be unable to work due to the many hours per week spent on dialysis and travelling between their homes and treatment centres. Their doctors were more likely to support their applications for government social grants (“disability grants”). This discrepancy in the access to social grants was a major concern for the PD patients.
The dialysis staff encouragement domain had the highest scores for both groups of patients, although PD patients had the higher scores. Dialysis staff were described as being like family and this support helped improve their quality of life. The support of their families, fellow patients and also government support was verbalised as being extremely important. Social support exerts a strong independent influence on subjective and objective QOL [
47]. PD patients called for the establishment of a patient support group and better financial support.
Reid et al. [
48] conducted a systematic review to synthesize the experiences of patients on haemodialysis and identified four broad themes which emerged from the included studies: a new dialysis-dependent self, a restricted life, regaining control, and relationships with health professionals. Our findings can be mapped to all of these themes which could be used as a framework to consider interventions for improving the experiences of patients on chronic dialysis. Of note is that in our study relatively little data emerged on regaining control, possibly indicating an area where we should focus our own interventions.
The findings of the study have caused us to review several aspects of our management of patients on chronic dialysis. In view of our strict PD-first policy, it is especially important that we do more for our PD patients to optimise their dialysis, facilitate their social functioning and improve the support that they receive. This includes educating them that doing occasional exchanges early or late when they have important social events will not cause long-term problems. It also includes more readily motivating for government grants for those who are not well enough to work or cannot find work, having our social worker increase her efforts to enlist the support of employers and arranging regular meetings of support groups which include both HD and PD patients.
We have recognised that the increased symptom burden and the sub-optimal laboratory data of our PD patients may be due, at least in part, to chronic under-dialysis and high rates of peritonitis. This has prompted us to start a quality improvement project which will be focused on reducing peritonitis rates, monitoring the adequacy of dialysis and improving anaemia management. Expedited transplantation or a switch to haemodialysis will be considered more readily for patients who are clearly not thriving on PD.
Our study has several limitations. It is a single-centre, cross-sectional study with relatively small numbers of patients. The experiences of our patients may not be the same as those of patients managed in well-resourced settings where there is broad access to RRT and where patient choice plays an important role in determining the dialysis modality. The findings we have reported may be especially relevant to RRT programmes in resource-limited settings where a PD-first policy is being followed and where the availability of haemodialysis is limited. Another potential limitation relates to the selection of our study participants. The patients in the focus groups were not selected randomly, but purposively so as to ensure the inclusion of a variety of participants, and opinions, in terms of sex, ethnicity, language and dialysis vintage. Having both males and females in a focus group, as was the case in this study, could potentially inhibit discussion about issues such as sexual dysfunction. However, our facilitator was struck by the open discussion around sexual function and did not get the impression that the participants were inhibited at all.
Acknowledgements
We thank the patients who generously gave of their time and allowed us to conduct and publish the study. We also thank Jonathan Maree for assistance with data collection.