Effect of Educational Intervention on Knowledge and Level of Adherence among Hemodialysis Patients: A Randomized Controlled Trial

Purpose The purpose of the study was to assess the impact of an educational intervention on the level of knowledge and adherence to the treatment regimen among hemodialysis (HD) patients as well as to describe the association between these variables. Methods In this randomized controlled trial, 160 HD patients at an HD centre of a 2030-bed tertiary teaching hospital in Southern India were randomly assigned into intervention (N = 80, received education and a booklet) and control (N = 80, received standard care) groups. Knowledge and adherence were measured preintervention and postintervention using a validated questionnaire for knowledge and the ESRD-AQ (End-Stage Renal Disease Questionnaire) for the level of adherence. The statistical analysis of the data was performed with the help of the Statistical Program SPSS version 19.0. The statistical significance level was set at 0.05. Results The increase in knowledge on disease management, fluid adherence, and dietary adherence in the intervention group was significantly higher compared to the control group. There was no significant correlation between knowledge and adherence. Adherence improved for all the domains, i.e., dialysis attendance, episodes of shortening, adherence to medication, fluid restriction, and dietary restriction. Adherence to fluid and dietary restriction was statistically significant. This trail is registered with https://clinicaltrials.gov/ct2/show/CTRI/2018/05/014166.


Introduction
Patients on dialysis experience assimilating complex treatment regimens, which includes monitoring blood glucose, intradialytic weight gain, bp monitoring, bill burden, physical activity, investigation routine, and adhering to treatment regimens. Patient education is not only a critical mechanism through which patients can have their questions, concerns, and needs regarding kidney disease care addressed but it is also a crucial pathway to ensure that patients can be taught to engage in self-management [1].
Nonadherence among HD patients includes the following, according to the National Kidney Foundation-Kidney Disease Outcomes Quality Initiative (NKF-KDOQI): (a) skipping or reducing the HD session; (b) consuming excessive amounts of potassium-and phosphorus-containing beverages and foods; and (c) failing to take medication as prescribed [2]. Nonadherence to dialysis treatment has been generally reported at rates between 8.5% and 22.1% worldwide [3]. Nonadherence is associated with increased mortality risk (skipping treatment, excessive IDWG, and high phosphate) and with hospitalization risk [4].
Patients' knowledge, either subjective or objective, does not seem to be sufcient. Hence, attention should be paid to supporting patients with more personalized knowledge [5] Some studies have shown that patient knowledge of disease and treatment is associated with an increased level of adherence [6][7][8]. Te success of treatment depends to a large extent on adherence to the strictly recommended therapeutic regimen. To improve adherence, patients' knowledge of disease management should be improved. Some studies have shown that patients' knowledge of disease and treatment is associated with an increased level of adherence [6,7].

Need for an Education Program and Rationale for the Current Study
Patients' understanding of hemodialysis and end-stage renal disease (ESRD) is essential for efective self-management and patient outcomes. Tere is a need for evidence and trials on the efect of therapeutic education among dialysis patients. Te available literature on therapeutic interventions of a focused nature has demonstrated positive efects, and evidence on the use of multidisciplinary care lacks certainty and majorly constitutes observational studies and nonrandomized controlled trials.

Design and Sample
A randomized controlled trial conducted from June 2017 to December 2020 was performed among 160 HD patients at a 2030-bed tertiary teaching hospital using block randomization and allocation concealment, and outcomes assessment was blinded. Te criteria for selecting the sample were as follows: (i) HD program two times a week; (ii) above 18 years of age; and (iii) ability to write, read, and understand the local language; patients with cognitive and psychologically diferent abilities and limited self-care were excluded. Te study population was randomly divided into two groups: the intervention group (received an educational intervention and a booklet) and the control group (standard care at the dialysis centre). Te educational intervention for each patient was administered for six months with reinforcement and addressing the patient's queries. Tis was followed up for one year. A postintervention assessment was conducted for knowledge at baseline, half-yearly, and end of 1 year. A postassessment at the end of one year was conducted for the quality of life, adherence level, and health service utilization.
Te baseline data demographics were collected with a proforma, knowledge of disease management was assessed with a selfadministered validated questionnaire and for measurement of the level of adherence, an ESRD-AQ was administered.

Sample Size Calculation
Knowledge was taken as the primary outcome variable for the computation of sample size. A mean diference of 0.5 SD (moderate efect size) is considered clinically important to consider that the intervention is efective. Assuming a power of 80% and a signifcance level of 5%, the sample size for the comparison of two group means is 64 per group. Adjusting for a 20% dropout rate, the required sample size per group is 80 per group ( Figure 1).

Randomization Procedure
Patients were randomized into two arms, i.e., control and intervention arms, using single block randomization. Block randomization with unequal block sizes of 4 and 6 is used to minimize selection bias, and 1 : 1 allocation will be done for intervention and control groups. For each block in the sequence, the permutation was selected using simple random sampling, and the same was followed for each of the 31 blocks, as shown in the example, and allocation concealment was done. Participants had an equal probability of being assigned to any given group.

Ethical Considerations
Ethical clearance (441/2015) was obtained from the Institutional Ethics Committee of Kasturba Hospital, MAHE, Manipal.
Phase 01 (Figures 2 and 3). Inputs from the KDOQI guidelines and expert' opinion on its adaptation and modifcation and cultural adaptation to the current population in the study were used to design the educational module. Major adaptations were in nutritional guidelines to the current local population. A judgmental validity was done. Judges who are professionals in the felds of nephrology, nutrition, pharmacy, and physical therapy evaluated this intervention guide. Te intervention guide was forwarded to the professionals listed above for feedback. Te PI discussed with the specialists the aim and goals of this stage. It was looked for ambiguity, such as ambiguous or badly phrased products, double-barrelled remarks, or jargon. For each item, the percentages of the entire agreement, agreement with small 2 Global Health, Epidemiology and Genomics   Global Health, Epidemiology and Genomics modifcations, agreement with large changes, and total disagreement were calculated. Any issue that received 70% or more total disagreement from the experts was removed from the teaching material. All the specialists were alerted to the elements that were in agreement with small and substantial adjustments. Many minor adjustments were fxed with their permission, and some important alterations were altered once a majority of experts agreed to that particular change. Te educational materials were written in English and converted to the local language with the back translation before the administration.
Phase 02. After the inclusion of the patients into one of the two groups, the following questionnaires were followed: (a) Patients were given a questionnaire on sociodemographic and clinical characteristics to characterize patients and identify their background information, and for the assessment of knowledge on disease management, a structured questionnaire. (b) Te ESRD-AQ explores all dimensions of HD patient adherence

Statistical Analysis
A mixed ANOVA (repeated measure) was performed to check if there was any signifcant diference in the average knowledge/ adherence across diferent time points as well as between the intervention group and the control group. A nonparametric approach of mixed ANOVA is performed using the R package "nparLD" to check if there was any signifcant diference in the average adherence/QoL across diferent time points as well as between the intervention group and control group as data violated the normality assumptions. As the outcome variable "Knowledge" was not normally distributed, quantile regression was used to determine the factors related to this outcome. As the outcome variable "Adherence" was not normally distributed, quantile regression was used to determine the factors related to this outcome. p < 0.05 is considered statistically signifcant, and analysis is performed using SPSS software.

Results
Te sample characteristics of the study population are described in Table 1. Participants included 80% males in the intervention group and 75.3% males in the control group. A higher proportion of participants in both groups had less than secondary education (52.2% and 43.8%) and were largely unemployed. Vintage of more than a year was comparably higher in both groups. Etiology-wise, hypertension predominated among those in the intervention group (41.3%), while in the control group, diabetes mellitus predominated (56.3%). All participants in the intervention group had comorbidities, while 7 (8.75%) in the control group did not. Tere was a statistically signifcant diference between the groups in the etiology of diseases (p � 0.001) and the presence of comorbidities (0.002). Regarding the mode of payment for treatment, cash payments were higher in both groups.
Tere was a signifcant increment in the knowledge after the intervention (p < 0.001) (Table 2). Similarly, also in the control group, there is a signifcant increment of knowledge score observed potentially due to data contamination and shift changes. It was observed from the between-groups comparison that there was a signifcant diference in the knowledge between the intervention group and the control group. Adherence to fuid and dietary restriction showed a signifcant improvement in the intervention group, while in the control group, there was a decrease in adherence (Table 3).Adherence for HD attendance, episodes of shortening, adherence to medication, and duration of shortening reported an increasing trend towards improvement; however, this trend is statistically insignifcant.
From Table 4, it is observed that the variable "Age" was a signifcant factor of knowledge. With every unit increase in age, the mean knowledge score reduced, and Table 5 shows individuals with cardiac morbidity and those with no other comorbidities, which were signifcant factors of adherence. Te mean adherence score is more in those who had cardiac comorbidity as compared to other comorbidities.

Discussion
Tis study reported that educational intervention can improve knowledge and adherence, by way of improving information, reinforcement, and limiting misconceptions about the disease. Knowledge of disease management and fuid and nutritional adherence improved signifcantly in the intervention group using an educational/cognitive intervention. Similar other studies using cognitive/educational intervention have improved knowledge on nutritional knowledge and binders, dietary phosphate intake, and weight gain control with a duration of intervention showing efect at minimum 2 months and 6 months with partial positive and positive benefts. Te benefts sustained beyond intervention were not reported in these studies [9][10][11][12].
Several studies [13,14] have also shown the positive impact of an educational session on knowledge levels. Ebrahimi et al. [15] also reported a signifcant increase in the level of their patients' knowledge concerning diet restrictions after an educational intervention. Similar results were found by other researchers [16] in a population of Iranian HD patients.
Tere was an improving trend in adherence in HD attendance, duration, and episodes of shortening of HD and a statistically signifcant improvement in fuid and dietary adherence. At baseline, the scores of the frst three domains of adherence were reported as high, and this could be attributed to the dialysis centre being the only tertiary-level hospital delivering high-quality care and the patient's perception of the quality of dialysis, while for fuid and dietary compliance, the scores were low due to a lack of knowledge, the myth about food to be avoided and eaten in moderation for their disease condition and the climatic condition making it difcult to adhere to fuid restriction. Te adherence outcome assessment was subjective, i.e., as the patient reported, and the educational intervention was  [10,[17][18][19][20][21]. Several other studies have used objectives such as biochemical parameter measurements, IDWG, and Kt/V with educational interventions for adherence outcome assessment with negative, positive, and partial positive effects [22][23][24][25][26]. Numerous studies also deployed behavioral interventions/counseling techniques to improve adherence.      [9,[27][28][29]. Tis study reported positive outcomes in adherence through subjective assessment among Indians, while many other similar studies and studies coupled with behavior and afective intervention and outcome assessed subjectively and objectively had diverse and heterogenous outcomes. Several studies have proven efective, partially efective, or negative results using educational, behavioral, or mixed interventions to improve patient adherence. Te current study used an educational intervention and patient-reported compliance to measure the efectiveness and found that an educational intervention can improve knowledge and adherence positively in Indian HD patients similar to previous authors who have demonstrated the importance of health literacy in health systems [30][31][32][33][34]. Te increase in knowledge level is not associated with increased adherence.

Limitation
Patient compliance was purely subjective in nature, and objective measurement of compliance was not performed, e.g., missed dialysis sessions, emergencies, and biochemical parameters. Tis study was limited to educational/cognitive interventions. Psychologic/afective interventions that appealed to the patient's feelings and emotions or social support and mixed interventions that involved a combination of the abovementioned intervention types were not tested. Te limitation is that the results cannot be generalized as the sample did not come from diferent regions of India and majorly constituted from coastal Karnataka. Moreover, the impact of factors such as noise, interruption by others, or participants' fatigue may infuence the answers of individuals. Tere was also a limitation of the time available to cover all thematic units, as the participants had only one educational session, which included a variety of thematic sections on CKD. For this reason and to enhance the educational outcome, the booklets were given to each participant separately after the intervention.

Conclusion
Tis study was a comprehensive approach and helped to improve the patient's knowledge of disease management and level of adherence. Tis education module can be used as a nurse-led intervention to improve patients' outcomes.

CKD:
Chronic kidney disease HD: Hemodialysis ESRD-AQ: End-Stage Renal Disease Questionnaire NKF-KDOQI: National Kidney Foundation-Kidney Disease Outcomes Quality Initiative IDWG: Interdialytic weight gain RRT: Renal replacement therapy ANOVA: Analysis of variance CBT: Cognitive behavior therapy.

Data Availability
Te data supporting the fndings of the current study are available from the corresponding author upon request.

Consent
Informed consents were obtained from all participants.

Disclosure
An earlier version of the manuscript has been presented in Research Square, available in https://www.researchsquare. com/article/rs-1194596/v1.

Conflicts of Interest
Te authors declare that there are no conficts of interest.

Authors' Contributions
BD provided intellectual input, conceived and designed the analysis, collected the data, was involved in methodology and analysis, developed the module, and revised the manuscript. RP and BU provided intellectual input, conceived and designed the analysis, were involved in the methodology and analysis, developed the module, and revised the manuscript. SB provided intellectual input and revised the manuscript. SC, VC, AS, and PM provided intellectual input, developed the module, and revised the manuscript.