Determinants Of Adherence To Dietary Prescription Among Patients With End-Stage Renal Disease Undergoing Hemodialysis In Dar Es Salaam, Tanzania


 Background: The increasing prevalence of chronic kidney disease means it has become a leading cause of death worldwide and a global health concern. The prevalence of chronic kidney disease in Tanzania is 12.4%. A major cause of mortality among patients with chronic kidney disease is non-adherence to dietary prescription. However, factors associated with non-adherence to dietary prescription remain unknown. Methods: This study used a quantitative descriptive cross sectional design, To investigate determinants of adherence to dietary prescription among patients with end-stage renal disease undergoing hemodialysis in Dar es Salaam Region, Tanzania. Data were collected using a self-administered structured questionnaire. Result: Data were obtained from 150 participants; 56.7% adhered to their prescribed diet. Factors significantly associated with adherence to dietary prescription were permanent residence in the city where the hospital was located (p=0.018), age (p=0.000), employment status (p=0.017), duration of chronic kidney disease (p=0.001), comorbidity (p=0.029), family awareness (p=0.003), affordability (p=0.044) and convenience of nutritional counseling services (p=0.046), and conversation with a healthcare provider regarding diet (p=0.039). Predictors of adherence were permanent residence (p=0.010), age (p=0.002), duration of chronic kidney disease (p=0.012), and conversation regarding diet with a healthcare provider (p=0.018). Conclusion These findings highlight a need for interventions to improve adherence to dietary prescription. Building understanding of determinants of adherence to dietary regimens among patients with end-stage renal disease is critical because of the poor progress associated with non-adherence. Further mixed method research should be conducted to gain an in-depth understanding of adherence to dietary prescription and associated factors.


Introduction
Chronic kidney disease (CKD) is a global health concern. The illness progresses to end-stage renal disease (ESRD), which is the advanced stage of CKD (1). Worldwide, CKD an accounted for an estimated 1.2 million deaths in 2005-2015, with most of these deaths in low-and middle-income countries (2). A systematic review conducted between 1995 and 2017 reported the prevalence of CKD in Africa was 10.1% (3). In sub Saharan Africa, the estimated CKD prevalence in the general population is 13.6% (4), that in East Africa is 14.4% (5), and that in rural Tanzania is 12.4% (6). However, the prevalence of CKD may vary with location. For example, Stanifer (7) reported a higher prevalence of CKD in urban compared with rural districts (15.2% vs. 7.0%) in Moshi municipality, in the Kilimanjaro Region of Tanzania.
Major causes of mortality among patients with ESRD are non-adherence to dietary prescription, hemodialysis sessions, and medication (8). Dietary modi cation is required to reduce the workload on kidney function and morbidity/mortality among patients on hemodialysis (9). Dietary recommendations aim to reduce further deterioration of kidney function, and require patients to limit food amounts to avoid accumulation of waste that cannot be excreted from the kidney (10). Complications associated with nonadherence to dietary prescription include bone demineralization, pulmonary edema, metabolic disorder, cardiovascular problems, and death (11). However, dietary adherence remains a challenge among patients with CKD, with a previous study reporting that 77% of patients deviated from their prescribed diet, predisposing them to further CKD complications (12).
In the African context, few studies have explored factors associated with dietary adherence. One study (13) reported factors associated with dietary adherence among adults with ESRD on hemodialysis were socioeconomic factors, condition and therapy factors, healthcare team-and system-related factors, patient-related factors, social gathering, and taste preferences.{Lambert, 2017 #30} Moreover, a study from Kenya that investigated dietary adherence among patients with CKD reported 36.3% adhered to their dietary prescription (14). Reasons for non-adherence identi ed in that study were lack of exibility of diet (80%), di culty in limiting uid intake (41.1%), and challenges in following the recommended diet (61.8%). However, that study collected subjective information from patients to establish these adherence factors (14). In addition, previous studies were conducted in contexts that may differ from Tanzania. For example, Tanzanian patients are likely to attend traditional and modern treatment, with the potential for non-adherence to either treatment modality (15,16).
The Tanzania Ministry of Health in collaboration with the private sector has invested in CKD management by training nephrology specialists, providing kidney transplant services, increasing the number of dialysis units, and reducing the expense of dialysis services (17). Despite these efforts, nutrition remains a challenge for patients with CKD. This is an important issue, as non-adherence to the prescribed diet may result in readmission and mortality due to metabolic disorder, anemia, pulmonary edema, or malnutrition. Dietary (including uids) modi cation has been shown to have positive outcomes among patients with ESRD (18). Therefore, this study aimed to investigate the determinants of adherence to dietary prescription among patients with ESRD undergoing hemodialysis in Dar es Salaam, Tanzania.

Study design and setting
We used a quantitative approach with a descriptive cross sectional design to determine the prevalence of adherence to dietary prescription and describe relationships between associated factors. This study was conducted at the renal unit (hemodialysis unit) in Muhimbili National Hospital, which is a tertiary hospital providing inpatient services (1200 bed capacity) and multispecialty clinic outpatient services that receive up to 2000 patients per day. Dar es Salaam is a large city (former capital city) located in the coastal region of Tanzania. This hospital was selected because it is the national referral hospital and university teaching hospital that provides hemodialysis services for many patients with ESRD. Approximately 80-100 patients (outpatients and inpatients) per day attend the nephrology department for hemodialysis.
Renal transplant procedure or operations are usually done once every 2 months.

Study population and selection criteria
The target population was patients with ESRD receiving hemodialysis; CKD accounts for around 80% of patients attending hemodialysis at the study hospital. All patients aged over 18 years that were diagnosed with ESRD (CKD stage 5) and had undergone hemodialysis for more than 1 month (i.e., attended the study hospital two or three times per week) were included in this study. Patients who were critically ill or unable to communicate were excluded at the time of data collection.

Sample size
The sample size of 150 patients was calculated using the Kish and Lislie formula (19). The proportion was derived from a study conducted in Kenya that found adherence to dietary prescription was 36.3%

Sampling technique
Systemic random sampling was used to ensure participants were representative of the study population. The registration book in the hemodialysis unit showed 100 patients attended hemodialysis each day; around 60 were patients with CKD. We aimed to study 20 patients per day. The formula used to calculate the sampling interval was: (k) = N/n, where K = 60/20=3. Therefore, the sampling interval was after every 3 patients.

Data collection
A self-administered paper-based questionnaire was used to collect data. This tool was adapted from a standardized questionnaire on dietary adherence and modi ed to address the objectives of this study based on a literature review (20). All questionnaire items were translated into Swahili as this is the common language used by participants. The questionnaire included closed-ended questions and covered major areas related to adherence to dietary recommendations e.g. in the past week, how many times did you follow dietary recommendation? Independent variables were demographic data, clinical characteristics, family support, and the healthcare system. The dependent variable was adherence to dietary prescription, which was measured by categorizing behaviors reported by participants. Patients who reported they adhered to dietary prescription "all of the time" or "most of the time" were classi ed as the adhered group and those who reported adherence "a few times," "very rarely," or "none of the time" formed the non-adhered group.

Pre-testing of the data collection tool
The questionnaire was tested for validity and reliability by conducting pilot study with 15 patients randomly selected from the hemodialysis unit. These participants were excluded at the time of actual data collection to avoid duplicated information. The Cronbach's alpha for the tool was 0.706. The results from the pilot study were used to modify the questionnaire as necessary before actual data collection.

Data collection and control
Participants were given a 21-item questionnaire that they were required to complete by themselves. The questionnaire took 20-30 minutes to complete, although some participants requested more time. Quality control was maintained by checking for completeness of the questionnaire or missing data. Data were stored securely in soft and hard copy, and only accessible to the researchers.

Data analysis and interpretation
Data were managed using SPSS version 25. The data were cleaned, coded, and then analyzed.
Descriptive statistics (means, standard deviations, percentages, and frequencies) were used to describe sociodemographic data, participants' clinical characteristics, and the prevalence of adherence, and presented in tables. Inferential statistics were also calculated, with chi-square tests used to test the association between independent categorical variables and adherence. Logistic regression was then used to predict the relationships between independent variables and the dependent variable. The signi cance level was set at 5% or p≤0.05.

Participants' sociodemographic characteristics
Of the 150 participants, 68.7% were male and 31.3% were female. The mean age was 48.4, SD±13.9 years (range 18-77 years). More than half of the participants were married (66.7%); 65.3% lived in Dar es Salaam and 34.7% lived upcountry i.e. outside Dar es Salaam region but with the boundary of Tanzania. The majority of participants were employed (including self-employed) (51.4%), and most had a secondary school education.
The rate of adherence to prescribed diets was 56.7%. Other demographic characteristics are presented in Table 1.

Participants' clinical characteristics
Participants' clinical information was categorized and analyzed using descriptive statistics.
The most common duration of CKD was 13-36 months (44.7%) and the most common duration of hemodialysis treatment was 13-36 months (43.3%). More than half of the participants had both diabetes and hypertension as comorbidities. The details are provided in Table 2. Other variables are shown in Table 3.  Table 4 below. One month ago 16 02 More than 1 month ago 23 15 When I started hemodialysis for the first time 29 26 Never 08 17 Other reasons 03 02 3.5 Regression analysis of factors associated with adherence to dietary prescription among patients with CKD on hemodialysis.
Variables that were significant in the chi-square tests were analyzed using logistic regression to evaluate if they predicted the likelihood of adherence among participants. As presented in Table 5, dietary adherence was significantly and independently affected by age (p=0.002), duration of CKD (p=0.002), permanent residence (p=0.011), and conversation with a healthcare provider about recommended diet (p=0.034).  (13,14,(21)(22)(23). Developing understanding of these determinants is important to inform development of strategies for effective nutritional counseling to improve adherence to dietary prescription.
Participants' age was signi cantly associated with adherence to dietary prescription. Older patients were more likely to adhere than younger patients. Other studies suggested that young adults reported facing di culties complying with dietary advice (21). In addition, unavailability of the recommended foods may result in non-adherence, as patients tend to consume what is available and accessible in their community (14). Our nding revealed that adherence to dietary prescription was signi cantly affected by permanent residence; therefore we speculated that food diversity might be affected by location of individual residence.
Being unemployed increases the state of poverty, which may result in non-adherence to dietary recommendations among patients with CKD (24). This suggests that having a reliable income may facilitate adherence because of the ability to purchase the recommended foods. However, a previous study (13) reported that patients who were not working were more likely to adhere to a renal diet than employed patients.
Duration of CKD was associated with adherence to dietary prescription among our participants. Patients with short duration of illness tended to adhere to dietary recommendations more than those with a long duration of illness. A previous study suggested that patients with long duration of CKD may encounter challenges in managing a complex diet that they are required to practice for years (25). In addition, patients with comorbidity (e.g., diabetes) faced more di culties in complying with diet and treatment compared with those without comorbidity (26). This was consistent with our nding that comorbidity was associated with adherence to dietary prescription. However, another study (14) reported no association between comorbidity and adherence to dietary prescription. We speculated that differences in the study settings and cultural diversity might have in uenced the discrepancy in these results.
We also found a signi cant relationship between family support and adherence to dietary prescription.
Patients who had family support tend to be more likely to adhere to dietary prescription than those without family support (27). Moreover, the person involved in food preparation needs su cient knowledge about food composition and its complexity in preparation (22). From our nding we speculate that, family member being aware of dietary recommendation and having the knowledge on food preparation increases the likelihood of adherence to patients.
Affordability of nutritional counseling was associated with adherence to dietary prescription among our participants. This was consistent with previous reports that indicated patients who were able to attend nutritional counseling were more likely to adhere to dietary prescription (23), whereas lower family income increased the risk for non-adherence (28). These results suggested that the ability of a patient to pay for nutritional counseling services could impact dietary adherence.
Similarly, patients who had conversations with nutritionists and received frequent sensitization on dietary issues from health workers adhered to dietary recommendations more than those who did not have such conversations (29). In our study, convenience of nutritional counseling services and conversations with a healthcare provider on dietary recommendation were signi cantly associated with adherence to dietary prescriptions. We believe that knowledge and sensitization on why a prescribed diet is necessary may in uence dietary adherence.
Age, permanent residence, duration of CKD, and conversation with a healthcare provider on dietary recommendations were identi ed as predictors of adherence to dietary prescription in this study. A previous study (21) reported that adult patients perceived themselves as more vulnerable to poor health outcomes, and therefore tended to comply with medical instructions. Our nding that patients with short duration of CKD were more likely to adhere to dietary recommendations than those with long duration may be because patients experience di culty complying with the same diet for years (25). Furthermore, participants' permanent residence and conversation with their healthcare provider on dietary recommendations predicted the likelihood of adherence. This may be because available food variety is linked with individual residence, and nutritional education is considered effective in increasing patient adherence to dietary recommendations (14).

Study limitation
This study has some limitations, the study was conducted at tertiary hospital (Muhimbili National Hospital); therefore, the ndings may differ from those in other settings. However, Muhimbili Hospital may be representative of different locations around Tanzania as it is a national referral hospital. In addition, this study used a cross sectional design, and does not provide evidence to establish causeeffect relationships. However, we revealed factors associated with dietary adherence among patients with CKD undergoing hemodialysis. This study required participants to report adherence behavior that occurred in the past, and might have led to recall bias. To minimize this issue, we used a questionnaire with items that explored easily memorable, general aspects of adherence. Moreover, we did not investigate adherence to speci c dietary recommendations (e.g., low protein, phosphorus, and potassium); therefore, the ndings showed general adherence to any dietary advice given to a participant by their healthcare providers.

Conclusion
This study demonstrated determinants of adherence to dietary prescription among patients with ESRD in Tanzania. Almost half of the study participants did not adhere to their prescribed diet, which highlights the need for interventions focused on the identi ed determinants to improve adherence to dietary prescription. This can be achieved by understanding barriers and modifying the healthcare delivery system to meet individual patient needs. Further research should be conducted to explore knowledge among patients with ESRD regarding dietary recommendations. Use of qualitative or mixed method designs will offer an in-depth understanding on the determinants of adherence to dietary prescription.
Patient in hemodialysis requires dietary modi cation to reduce severity of morbidity and mortality. This study highlights the key factors associated with adherence to dietary prescription among patient with end stage renal disease. Also, it provides information on the role of adherence to nutritional therapy in managing ESRD. Furthermore adds knowledge to healthcare providers on predictive factors toward dietary adherence among patient with ESRD.

Policy implication
Provide information to policy maker on dietary adherence and related strategies on development or review existing policy on management of CKD. The current study highlights the key areas on dietary adherence that healthcare provider can emphasize on provision of nutritional management on patient with CKD.

Declarations Ethical Considerations
Aga Khan University Ethical Review Committee approved this study. Permission to conduct the study was obtained from the management of Muhimbili National Hospital. Written informed consent was obtained from each participant before the study started. Consent forms were written in Kiswahili to ensure participants fully understood. The consent form included an explanation of the purpose of the study, con dentially of the information, and the right to withdraw at any point.

Consent for publication
Not applicable.

Availability of data and materials
The dataset generated and/or analyzed during the present study is available from the corresponding author upon reasonable request. Permission to access this data will be required from the Aga Khan University.