This study was conducted across five private healthcare centres in the cities of Bloemfontein, Kroonstad, and Kimberley, in the Republic of South Africa. The private facilities in the Free State province included Pelonomi, Universitas, Bram Fischer, and Kroonstad. In the Northern Cape province, the private study site was Kimberley. Of the 120 patients seen at these facilities, only 78 met the inclusion criteria for the study. The study was conducted only in private settings, as public hospitals did not grant approval, citing concerns about research burden on dialysis patients and prioritising internal staff research projects.

This study adopted a quantitative, cross-sectional, prospective analytic research approach to explore the correlation between depression levels and dialysis adequacy in CKD patients. This design was chosen because it enables observation and analysis of relationships between variables over time, ensuring a robust understanding of the interplay between mental health and dialysis outcomes (American Journal Experts 2022). The study is more accurately described as a quantitative cross-sectional analytic study with prospective data collection. The cross-sectional component refers to the measurement of both depression and haemodialysis adequacy at a single point in time for each participant, which is appropriate for assessing their correlation, as cross-sectional studies provide a snapshot of variables within a population at a specific time (Thomas 2023). The prospective aspect relates to the method of data collection, whereby respondents are recruited and assessed in real time as they present for haemodialysis, rather than relying on retrospective records. Thus, the study does not involve follow-up over time but prospectively collects cross-sectional data, ensuring standardised measurement and improved data quality, as cross-sectional studies involve the simultaneous measurement of exposure and outcome without longitudinal follow-up (Prabhakar & Kaushal 2024).

In determining an appropriate sample size, we considered the total number of patients at each study site and the study’s inclusion and exclusion criteria. A total population sampling approach was employed.

This approach was suitable given the small, well-defined, and accessible population, reducing sampling error.

However, selection bias is acknowledged, as findings from private clinics may not be fully generalisable to public-sector patients due to differences in socioeconomic status, access to care, and health system resources.

The study evaluated the entire accessible population of 120 patients across five centres. Following the systematic application of inclusion and exclusion criteria, all 78 eligible individuals were included in the study to ensure a homogeneous group and maximise statistical power, as confirmed through consultation with a biostatistician. This approach aligned with recommendations that sampling strategies in quantitative research should closely match the study population and objectives to ensure the validity and reliability of findings (Creswell 2023).

Respondents were recruited from the selected healthcare centres through collaboration with healthcare professionals and patient care teams. Recruitment strategies included direct engagement with potential respondents and the distribution of study information sheets. The principal investigator introduced the study’s concept to all patients at each location, encouraging both patients and nursing staff to actively participate and freely express their views and questions relating to the study. After satisfactorily addressing all inquiries, the study team, in partnership with the nursing staff, identified patients who expressed interest in participating in the study.

The inclusion and exclusion criteria were explained in detail to the willing respondents to ensure clarity and understanding. Subsequently, an informed consent form and a study information sheet were distributed to the willing respondents. After the consultation sessions concluded and respondents expressed their satisfaction, they signed a consent form.

Data collection was conducted over a 3-month period (June 2024 to August 2024). Respondents attended three haemodialysis sessions per week, resulting in a total of 12 sessions per month and 36 sessions over the study period.

Measurement of dialysis adequacy:

Depression assessment: Respondents completed the BDI questionnaire at the start of the study. The questionnaire was administered once and analysed in consultation with a qualified psychologist.

This study employed statistical analysis system (SAS) software (Cary, NC, United States) to analyse the collected data. Data were captured in a Microsoft Excel spreadsheet (Redmond, WA, United States) by the principal investigator, screened for completeness, cleaned for missing values, and quality checked before submission to a biostatistician for analysis. The Shapiro–Wilk test was used to assess the normality of numerical variables. Categorical variables were summarised using frequencies and percentages, and group comparisons were performed using the chi-square test. Normally distributed continuous variables were compared using Student’s t-test and analysis of variance (ANOVA), while non-normally distributed variables were analysed using the Mann–Whitney U test and Kruskal–Wallis test. Chicco, Sichenze and Jurman (2025) explain that statistics provides researchers with a range of valuable analytical tools, including several univariate tests designed to examine relationships between numerical variables. These include Student’s t-test, Mann–Whitney U test, chi-squared test, and Kruskal–Wallis test. The authors further note that such tests yield various statistical outputs, particularly p-values, which quantify the strength of evidence against the null hypothesis and inform decisions to either reject or accept it based on a predetermined significance level. Correlation analysis and logistic regression were conducted to examine the association between mental health and haemodialysis adequacy. A significance level of α = 0.05 was applied to all statistical analyses to address the study objectives comprehensively.

The study ensured validity and reliability through several methodological strategies. Firstly, rigorous inclusion and exclusion criteria were applied to ensure a well-defined and homogeneous study population, thereby enhancing internal validity and reducing selection bias. Secondly, standardised and validated measurement tools were used, including the BDI tool for assessing depressive symptoms and Kt/V calculations for determining haemodialysis adequacy, both of which are widely recognised for their reliability and clinical validity. In addition, data collection procedures were standardised across all respondents to ensure consistency in measurement and reduce inter-observer variability. Furthermore, appropriate statistical techniques were employed to enhance the accuracy, reliability, and reproducibility of the findings, including methods to control for potential confounding variables where applicable. Collectively, these measures strengthened the scientific rigour of the study and ensured that the findings are both credible and generalisable within the study context.

Ethical considerations were of the utmost importance in this study. Ethical clearance was obtained from the University of the Free State Health Sciences Research Ethics Committee (ethical clearance number: UFS-HSD2023/2414/3004). Additionally, ethical approval for the study was obtained from the private company’s research committee prior to the commencement of the research. Respondents received an information sheet detailing the study’s objectives, expectations, duration, and investigator contact information. Informed consent was obtained, and respondents were informed of their right to withdraw from the study at any time without facing negative consequences.

Pseudonymised data were used to protect respondents’ confidentiality. De-identification techniques, such as pseudonymisation and anonymisation, play an important role in facilitating such secondary uses and disclosures of data (Thaldar 2023). In this study, data were securely stored in the university’s digital storage system, with physical documents, including questionnaires and chronic dialysis sheets, locked in a safe.

Portable devices used for data collection were encrypted, and hard copies containing identifiable information were securely stored and accessible only to authorised research team members. An information sheet detailing the study’s objectives, expectations, duration, and the investigator’s contact information was provided to each respondent. All willing respondents gave informed consent and were then informed that they had the right to withdraw from the study at any time without facing negative consequences. No respondents withdrew from the study.

Of the 78 respondents in this study, 45 were female, and 33 were male. The mean age of the respondents was 50 years. Ten respondents used wheelchairs, and one was visually impaired and was assisted by nursing staff to complete the questionnaire; the nursing staff read the questions and allowed the respondent to choose the answer. All 78 respondents had some level of formal education.

Dialysis adequacy was evaluated using daily and average monthly online Kt/V values, calculated from clinical data collected during the study period; the findings are summarised in Table 1. It presents the distribution of key clinical variables, such as Kt/V values, using medians and interquartile ranges, which are appropriate for skewed biomedical data. These measures provide a baseline understanding of variability in dialysis adequacy, supporting accurate interpretation of treatment effectiveness and the identification of clinically meaningful differences among respondents.

The average Kt/V for month 1 and month 2 remained consistent, with a median of 1.03. However, the month 2 recorded a maximum Kt/V value of 1.76, compared with the month 1 maximum Kt/V of 1.24 and a minimum of 0.68. The overall median Kt/V over the 3 months was 1.04, with a maximum Kt/V value of 1.34 and an interquartile range of 1.00 to 1.10. This indicates that Kt/V values remained stable over the 3 months, with an overall median of 1.04, below the minimum prescribed target of 1.2. Most patients, therefore, did not achieve the recommended dialysis adequacy, although a few individuals reached higher Kt/V values, as reflected in the maximum measurements.

Figure 1, representing of month 3 from 01 August 2024 to 30 August 2024, demonstrates that the first treatment engaged four respondents, the second treatment involved nine respondents, and the concluding third treatment comprised six respondents who successfully reached the Kt/V target of 1.4.

Creatinine and urea clearance levels were assessed monthly from routinely collected blood samples to further measure dialysis adequacy. Of the 78 respondents, 75 participated in urea testing and 76 in creatinine testing. The outcomes of the clearance assessment are presented in Table 2. It presents detailed descriptive statistics for urea and creatinine, two critical markers of dialysis effectiveness.

By comparing central tendency and variability, the Table 2 helps assess how well high-flux haemodialysis is clearing metabolic waste, thereby serving as an indirect indicator of treatment performance.

The median urea level over 3 months was 20.13 mmol/L, with an interquartile range of 16.77 mmol/L – 22.40 mmol/L.

A maximum urea level of 41.30 mmol/L was recorded in month 2, surpassing the overall maximum of 39.73 mmol/L, as presented in Table 2. Month 1 recorded the highest mean creatinine value of 791.84μmol/L, along with a peak of 1504.00 μmol/L, surpassing all subsequent months. The lowest creatinine level, 122.00 μmol/L, was noted in month 3. Additionally, month 3 and the ‘overall creatinine’ category both exhibited a maximum creatinine value of 1448.00 μmol/L. The overall creatinine’s mean was determined to be 778.30 μmol/L, with a standard deviation of 300.20 μmol/L.

The respondents completed the BDI questionnaire at the start of the study. The questionnaire was administered once and analysed in consultation with a qualified psychologist. It is a self-report tool completed by every respondent privately at the dialysis centre. The questionnaire was completed anonymously; however, each respondent was assigned a unique identification number for tracking purposes. The questionnaire was distributed to the respondents and collected on the same day once they indicated that they had finished. A few respondents requested to take the questionnaire home to complete and return it at their next dialysis session. The confidentiality declaration was signed by a psychologist who is part of the research team. The BDI questionnaire was utilised to evaluate the severity of depression. Respondents rated their agreement with the 21 items on a 4-point Likert scale, where scores ranged from 0 (‘Never’) to 3 (‘Always’). Based on their scores, the respondents were classified into categories defined by the BDI scoring system. Individuals are classified into six categories, with the first, scores under 10, labelled as ‘Considered Normal’, and the last ‘Extreme’, according to BDI criteria.

The BDI uses a straightforward scoring system:

Table 3 shows the frequency and percentage distribution of responses to individual BDI items. It highlights the prevalence and severity of depressive symptoms among respondents, providing insight into the psychosocial burden associated with chronic dialysis and its potential interaction with treatment adequacy. Respondents rated 21 items on a 4-point Likert scale ranging from 0 (‘Never’) to 3 (‘Always’).

Total BDI scores were calculated by summing individual item responses, and the frequency distribution for each item was analysed.

The median BDI score was 17 (range: 1–50), as presented in Table 4. A median BDI score of 17 indicates that the majority of respondents exhibited depressive symptoms. According to the BDI scoring system, a total score between 17 and 20 falls within the range for borderline clinical depression. Table 4 summarises overall depression severity using total BDI scores. The distribution of scores enables assessment of central tendency and spread, helping to determine the overall mental health status of the study population and its potential relationship with clinical outcomes.

Table 5 classifies respondents into standard BDI severity categories and shows their frequency distribution. It is important because it translates raw scores into clinically meaningful groupings, enabling easier interpretation of depression prevalence and facilitating comparisons with measures of dialysis adequacy. Respondents were categorised based on total BDI scores, and the categories were composed as follows:

Table 6 presents Spearman’s correlation coefficients assessing the relationships between depression scores and clinical variables. It is crucial for identifying potential associations between psychological well-being and dialysis effectiveness, especially where both physiological and psychosocial factors may interact. The statistical analysis was conducted using the Spearman’s correlation coefficient. For this purpose, the BDI total score and overall and Kt/V average are presented in Table 6.

A significant negative correlation was observed between Kt/V and BDI scores (r = −0.257, p = 0.0231), indicating that lower Kt/V values are associated with higher depression levels. Since r² = 0.0661, only 6.61% of the change in BDI score is due to the change in overall Kt/V, where 93.39% of the change in BDI score is because of other factors. The BDI and overall urea: The analysis showed a non-significant correlation, with p = 0.779, r = −0.0329, and r² = 0.0011. The BDI and overall creatinine: A non-significant correlation was observed between the two variables, p = 0.654 and r = 0.0522, with r² = 0.0027. A weak but statistically significant inverse relationship was observed between Kt/V and depression scores, indicating that lower dialysis adequacy was associated with higher levels of depressive symptoms. No significant associations were found between depression and urea or creatinine levels.

The correlation analysis between the overall Kt/V and BDI total scores is reported in Table 7. It helps determine whether psychological status is associated with differences in treatment effectiveness, which may suggest clinically relevant links between mental health and dialysis outcomes. The categories of mild mood disturbance and borderline clinical depression each comprised 20 respondents. Both the mild mood disturbance and moderate depression categories reached identical maximum Kt/V values of 1.34 and the same median value of 1.03; however, their interquartile ranges differ.

Table 8 reports Cronbach’s alpha (α) values, indicating the internal consistency of the BDI used in the study. Demonstrating strong reliability supports confidence in the measurement of depression and strengthens the validity of subsequent analyses. The BDI demonstrated good internal consistency as shown in Table 8, with a Cronbach’s α of 0.886, confirming the reliability of the instrument.

This study was geographically limited to the Northern Cape and Free State provinces, which may affect generalisability. The final sample of 78 respondents, below the intended target, may reduce statistical power.

Dialysis adequacy was assessed solely via OCM, without validation against alternative methods or consideration of residual renal function. Important clinical parameters, such as body mass index (BMI), haematocrit, ultrafiltration volume, and laboratory-based urea kinetics, were not captured, potentially limiting the depth and interpretability of findings. Future studies should incorporate broader populations, larger sample sizes, and multimodal adequacy measures to enhance validity and applicability. Furthermore, all haemodialysis machines utilised are of the same brand and model; thus, it is recommended to evaluate other machines to ensure the reliability of the results. Online measurement is used as a single marker of adequacy; renal residual function is not considered, pathology is calculated not recorded, BMI, hematocrit and ultrafiltration not included.

Another limitation of this study is that depression was assessed using the BDI tool without a formal clinical diagnostic interview. The BDI measures depressive symptom severity rather than providing a clinical diagnosis of depression, which may introduce misclassification bias. Consequently, some respondents may have been incorrectly classified regarding depression status, which should be considered when interpreting the findings.

A limitation of this study is the limited consideration of the cultural and linguistic validity of the BDI in the South African context. Although widely validated, its use in multilingual and diverse populations may be affected by language interpretation and literacy levels. Respondents with language, literacy, or visual difficulties were assisted by the principal investigator and nursing staff through verbal translation and clarification in English and relevant local languages, thereby improving accessibility but potentially introducing response bias.