Recruitment and participant baseline characteristics in the dialysis outcomes in those aged 65 years or older study

The study has an accelerated prospective cohort longitudinal design, comprised of cross-sectional and longitudinal components (Fig. 1). The methods have been described previously [2]. In particular, individuals who had a very limited life expectancy or significant active co-morbidity which was affecting their health were excluded by principal investigator at each site, as this would clearly impact on the perceived HRQoL questionnaires, but may not be attributable to dialysis and ESKD specifically. Recruitment was targeted to try and get at least one set of interviews at 12 months after enrolment for all participants where possible.

Eligible patients resided in one of three District Health Board (DHB) regions within NZ (Counties Manukau, Hawkes Bay or Southern) which reflect the diversity of the New Zealand population with Counties Manukau, a large urban population with high representation of Māori and Pacific people; Hawkes Bay, a regional farming centre which also has a high proportion of Māori and the Southern region with a predominantly European population spread over a large demographic area. All participants were 65 years or older, their clinical team considered them well enough to be approached to participate, and were either established on dialysis at the baseline interview or had a eGFR 15 ml/min 1.73 m² or less and had commenced pre-dialysis education. In practice, this meant that some patients were already on dialysis for a number of years at the time of recruitment, whereas other patients had just commenced treatment. Clinicians from each DHB approached eligible patients and invited them to participate in the study. Ethical approval was granted by the New Zealand Health and Disability Ethics Committee (MEC/10/08/084) and the study is registered with Australian and New Zealand Clinical Trials Registry (ACTRN12611000024943).

Following consent, the patients were telephoned by a trained interviewer to arrange an interview either face to face or by telephone depending on the participants’ preference. Interview questions were read aloud to participants, who were also provided with a printed sheet of response options to facilitate the interviews. The baseline interview was predominantly comprised of structured closed-response questions (see below).

Clinical data was collected at the same time as the first (baseline) interview for each participant. Clinicians completed a standardised clinical data abstraction form based on report forms used for the Dialysis Outcomes and Practice Patterns Study. Data collected included ethnicity, age, sex, co-morbidities, laboratory data, anthropometric measures, prescribed medications and details of renal replacement therapy [3] In addition, we obtained aggregate data from routinely collected clinical and administrative hospital datasets included in the Australian and New Zealand Dialysis and Transplantation Registry (ANZDATA) to compare characteristics of our cohort with the broader New Zealand wide dialysis population in this age group [4].

We collected socio-demographic characteristics using the same questions used in the 2009 New Zealand census: sex, age, relationship status, occupation, living arrangements and self-reported ethnicity [5]. As participants were able to select multiple ethnicities, ethnicity was prioritised into Māori (New Zealand’s indigenous population), Pacific peoples (Samoan, Tongan, Fijian, Cook Islander), other ethnicities (e.g. other European, Chinese, Indian) or New Zealand European [5, 6]. Socioeconomic data included home ownership [5], whether income was solely dependent on Government Superannuation [7], use of a means-tested community services card (to provide additional subsided health care) [8], self-reported standard of living (responses: high, fairly high, medium, fairly low, low) [9], adequacy of household income to meet every day needs (not enough, just enough, enough or more than enough), and education (the responses were merged into 3 categories: none, secondary (age 12–18 years), post-secondary) [9].

The New Zealand Deprivation Index (NZDep2006) is calculated for small New Zealand geographical areas using 9 dimensions of socio-economic disadvantage to create a summary score to estimate socioeconomic status using data collected in Statistics New Zealand’s 2006 Census of Population and Dwellings [10]. The higher the number (range 1–10) the greater the estimated socioeconomic deprivation.

General health status was assessed using the EQ-5D-3 L [11], as previously used in studies of Chronic Kidney Disease (CKD) populations [11, 12]. The Kidney Disease Quality of Life - 36 (KDQoL-36) was used to assess the functioning and well-being of people with CKD [12‐14].

We collected responses to the Functional Assessment of Chronic Illness Therapy (FACIT) survey, which is a collection of health-related quality of life measures targeted to the management of chronic disease and has been validated in elderly populations [15]. The higher the score the higher the level of satisfaction with health care. A total FACIT score was calculated as a simple (unweighted) sum of all the subscale scores, rescaled to a 0 to 100 scale. This was calculated only for people with valid sub-scales for all dimensions.

The level of disability for participants was measured using the WHODAS II 12-item measure which assesses activity limitations and participation over the past 30 days using a five point Likert scale (“none/mild/moderate/extreme/cannot do”) for each item. [16] A WHODAS disability score was derived using the summed approach (where 0 = no disability and 48 = disability). When one WHODAS item was missing, the individual’s average was imputed; when more items were missing, the disability score was not derived. [16]

Well-being was measured using the Personal Wellbeing Index [17]. This scale is an eight-item measure assessing the level of satisfaction with; standard of living, health, achieving in life, relationships, safety, community connectedness, future security and spirituality or religion. Respondents rated their satisfaction on an 11 point scale (0 completely dissatisfied to 10 completely satisfied). Scores for individuals were derived as an average across all items and converted to a zero to 100 scale.

Between 1 January 2010 and 31 March 2014 there were 388 potential participants of whom 56 were ineligible for the first interview because; they died before interview was scheduled (n = 11) or a clinician advised interview would be inappropriate given the poor health status of the patient or anticipated limited life expectancy (n = 45). Of the 332 potential participants, 107 (32%) declined to participate in the study, leaving 225 (68%) participants (Figs. 1 & 2). The demographic characteristics of patients who agreed and declined to participate in the study is provided in Table 1. There were a number of statistically significant differences between the two groups. Participants were more likely to be male, of European ethnicity and on dialysis. Over half of the 225 participants (54%) completed a face to face interview, 45% by telephone and 1 completed a postal questionnaire. Interviews took on average one hour to complete. 25 participants (11%) had limited or no English language ability and were assisted by a bi-lingual translator.

A comparison of DOS65’s population with the entire comparative age group on dialysis in NZ [4], found there was no significant difference in the distribution according to age or sex. In DOS65, 42% were dialysing at home which reflects national trends [4]. When compared to the New Zealand wide dialysis population of the same age [4], DOS65 had a higher proportion of Pacific participants (23% vs 17%), a similar proportion of Māori participants (22% vs 25%) and a slightly lower proportion of New Zealand Europeans (43% vs 49%). When compared to the general NZ population over the age of 65, Māori (22% vs 5%) and Pacific people (23% vs 2%) were over represented in the study population [9].

Participant characteristics are summarised in Table 1. Nearly two thirds of the study cohort were male and the median age at interview was 71 years (interquartile range, IQR 68 to 76 years). Three quarters (n = 169) of the patients were receiving dialysis (109 haemodialysis and 60 peritoneal dialysis). Of those on dialysis, 70 (41%) were dialysing at home; and 99 (59%) underwent centre-based haemodialysis. 119 (70%) had been dialysing for more than one year at baseline; 57 (34%) had been dialysing for between 1 and 3 years and 62 (37%) had been dialysing for more than 3 years. Fifty (30%) were incident dialysis patients (started dialysis within the past 12 months). Of the 56 non-dialysis patients, 9 had elected for conservative treatment (i.e. had made a conscious choice to not go onto dialysis), and 47 were defined as pre-dialysis. There was no sub-cohort effect observed across dialysis duration (Table 2). Of the pre-dialysis patients, 5 had been deemed medically unfit to start dialysis, 41 had chosen to dialyse but had not reached the point at which dialysis needed to be started, and 1 patient had not yet made up his/her mind about whether to dialyse or not.

Diabetes mellitus was the most common cause of ESKD (41%) followed by glomerulonephritis (17%) and hypertensive disease (17%). The majority of participants had 3 or more comorbidities, of which cardiovascular disease (CVD) (67%), followed by diabetes mellitus (53%) were the most common. The number of comorbidities ranged from zero to 6 with a median of 3 (IQR 2 to 4) (Table 1). Over half the population had previously smoked (52%) though there were very few current smokers (6%). The population was less well educated than the general NZ population of the same age with 48% of the study population not having achieved a secondary school education as compared with 39% of the NZ population over the age of 65 (2013 NZ Census). The study population lived in more deprived areas than the general NZ population over the age of 65. DOS65 participants were poorer socioeconomically, according to the NZDep, than the general population (6.9 and 5.4 respectively, p < 0.01), and more than half (54%) lived in areas classified as having the poorest three deciles. (Personal communication Dr. June Atkinson Department of Public Health, University of Otago Wellington, NZ; data sourced from Statistics New Zealand).

Table 3 describes participants’ sociodemographic characteristics by baseline renal replacement therapy, dialysis duration and location of treatment. The length of time on dialysis (dialysis vintage) did not vary by gender, age, ethnicity, distance from dialysis facility or any makers of socioeconomic status. In contrast dialysis modality and location was highly associated with socio-demographic factors. Female participants were less likely to dialyse at home than males (27 vs 49%). When compared to the NZ European population a smaller proportion of Māori (25 vs 61%) or Pacific people (24vs 61%) dialysed at home. Participants living further away from the dialysis centre were more likely to undertake a home therapy, particularly peritoneal dialysis. Patients dialysing in-centre (compared to those dialysing at home), were more likely to be socioeconomically disadvantaged as indicated by not owning their own home (72 vs 48%), being more likely to have a community services card (85%vs 62%), reporting a low/very low standard of living (14% vs 3%) and living in a more deprived NZDep area (mean 7.6 vs 6.1). Age had no impact on dialysis location and people living alone were no less likely to undertake a home therapy. Patients dialysing at home had a similar number of comorbidities to those dialysing in-centre. The characteristics of the pre-dialysis population was similar to the population on peritoneal dialysis. Females, Māori and Pacific people, those who lived closer to a dialysis centre, and those who were socioeconomically poorer, were more likely to undertake haemodialysis when compared to the peritoneal dialysis or predialysis population.

Table 4 presents self-reported HR-QOL, disability and personal well-being measures at baseline. With the exception of dialysis location on EQ-5D-3 L VAS, these self-reported scales of health and disability were not impacted by being on dialysis or by dialysis modality, location or duration.