Trial Overview: The overall design of TEACH-VET is depicted in Figure 1. First, the study uses an electronic health records (EHR)–based strategy to identify veterans with advanced CKD (source cohort) and assess their status/need of ongoing specialty nephrology care and CPE. Then, in a mixed method randomized controlled trial (RCT), the study plans to enroll 544 Veterans from the source cohort, and randomize them in 1:1 allocation to the CPE/intervention arm vs. usual clinical care supplemented by kidney disease education material, enhanced usual care (EUC) arm. The study aims to compare the effects of intervention/control on parameters of informed dialysis decision and dialysis modality selection, dialysis modality use, and several clinical, patient-centered and health services outcomes post-ESKD. The also has a qualitative component, which uses semi-structured interviews to explore Veterans’ perceived satisfaction with CPE, their preferences for face-to-face or tele-CPE, and their perceived barriers and facilitators in the selection and use of preferred dialysis modality. The quantitative and qualitative data are collected and will be analyzed separately, and the results will be integrated for a more comprehensive understanding.
Conceptual Framework: The structure of TEACH-VET is based on the modified Bandura’s model of social cognitive theory.[14] Social cognitive theory identifies a set of core determinants including knowledge of health risks and health benefits from different health practices, self-efficacy, outcome expectations, and perceived facilitators and social and structural impediments to the changes they seek. These core determinants create the preconditions for, and play a central role in, human motivation, action, and health decisions. DeWalt et al.[15] successfully modified this model in a randomized evaluation of educational intervention in heart failure patients arguing that patient action, i.e., informed decision-making in TEACH-VET leads to positive changes in patient health-related outcomes.
Hypothesis and Rationale: We hypothesize that a system-based application of universal CPE and patient-centered initiation of renal replacement therapy will increase home dialysis utilization and improve multiple clinical, patient-centered and health service outcomes (Figure 2). Specifically, CPE will increase Veterans’ self-efficacy, i.e., knowledge of CKD and its management so that they become more confident in making an informed choice for their disease management and dialysis treatment. The contention is also that Veterans’ behavior post-CPE will lead to increased use of home dialysis compared to the usual care group (primary outcome). According to social cognitive theory,[15] individual person-level determinants (e.g., knowledge and confidence) may increase the likelihood of an individual’s executing a behavior (e.g., informed decision-making and self-management). Additionally, environmental factors can also influence behavior; as such, environmental factors the Veterans perceive as barriers and facilitators will be examined (qualitative phase). Finally, CPE-induced behavioral changes may show positive impact on post-ESKD outcomes.
Study Population: TEACH-VET was launched in August 2020 across the North Florida/South Georgia (NF/SG) Veterans Health System (VHS), one of the busiest VHS in the US comprising 2 VA medical centers and 10 outpatient clinics. Based on the recommendations by the professional renal organizations and CMS, TEACH-VET aims to target all adult (> 18 years old) Veterans with advanced stage 4 and 5 CKD not on dialysis for enrollment. The study excludes Veterans who are non-English-speaking, homeless or living in assisted living facilities, and with dementia or less than 6-months life-expectancy.
Enrollment strategy: To ensure the enrollment targets all prevalent advanced CKD patients, in addition to directly approaching the Veterans attending the renal clinic, TEACH-VET recruits participants through our recently published, EHR–based ‘Opt-Out Source Cohort Strategy.[16] In brief, a ‘source cohort’ of all actively registered Veterans at NF/SG VHS with ICD-10 codes for stage 4(N18.4) and 5 CKD(N18.5) or two latest outpatient estimated glomerular filtration rate (eGFR) of less than 30ml/min at least 90-days apart is generated. The cohort is then sorted in a random order, and the potential participants are approached in consecutive order for their status/need for specialty nephrology care and CPE. All eligible and interested participants are then enrolled into the second phase RCT.
Baseline data collection and Randomization: Once enrolled, all participants provide baseline data comprising of patient-reported socio-demographics, education, household composition, and annual family income. Participants are assessed for health literacy by Rapid Estimate of Adult Literacy in Medicine-short form, medical comorbidity by the Charleston Comorbidity Index, and HRQoL by Kidney disease quality of life (KDQoL-36), excluding dialysis items.[17-19] CKD awareness is assessed by the prior validated instrument by Wright et al.[20] Considering this and other similar validated CKD knowledge instruments lack the domains of ESKD knowledge—essential for informed dialysis decision—the team has developed and pilot tested a 29-item ESKD knowledge questionnaire.[21] This questionnaire will be further refined during the TEACH-VET, and the team will report on its findings. A full list and timeline of all collected variables are listed in Figure 1 and Table 1.
All enrolled participants are randomized by a computer-generated block randomization schedule devised by the study statistician, in 1:1 ratio into CPE or EUC arm. Considering the primary outcome of dialysis modality use and the strong known influence of socioeconomic factors,[22] the randomization is stratified by the stage of CKD (4 or 5) and annual family income (250% above or below federal poverty level adjusted for total number of household members).[23]
Intervention/CPE arm: Participants and their preferred care partner(s) in the intervention arm receive a standardized, evidence-based, two-phase CPE by trained renal educators in an Intent-to-Teach manner. The protocol covers the domains of education recommended by the professional renal organizations and CMS (Table 2),[24, 25, 4, 5] with an interactive, instructor-led audio-visual education, followed by individual patient-oriented counseling session that includes lifestyle simulation discussions. Prior studies have shown the advantages of such two-phase approach on comprehension, fears, and home dialysis selection.[26, 21] Over last decade, we have tested, refined and validated this protocol at two geographically distinct universities and affiliated VAs within the US to ensure literacy level and cultural relevance for the target patient population.[11, 12] For this study, we further pilot-tested the intervention with a local Veteran Engagement Committee made up of a diverse group of 12 Veterans and Veteran caregiver volunteers from Florida. This committee provided specific feedback to further hone the language used and explanations given for describing kidney disease and its management to fellow Veterans. Recently, we demonstrated our protocol can be delivered either face-to-face or through telemedicine with equivalent outcomes in terms of confidence in dialysis decision-making and home dialysis selection.[21]
To ensure the intervention is standardized and uniform throughout the study, the renal educators are trained by licensed renal providers in the content, and by experienced patient educators in the delivery of the CPE prior to their involvement in the study. Additionally, with the participants’ permission, all CPE sessions are recorded for the first 3 months of the study or after initiation of the new educator, and 10% of randomly selected CPE are recorded throughout the study period. The recorded data is reviewed for credibility, competence, and thoroughness of the educator interactions during CPE. Finally, the study tracks the amount of time educators spend with each participant for individual counseling, reviews the fidelity of important pre-defined topics and their delivery, and keeps detailed notes of any deviations from the CPE protocol. Feedback and additional training is provided as needed to ensure uniformity and standardization. Patients having any question or concerns after education are provided the opportunity to discuss with a licensed dialysis nurse or provider proficient in all renal replacement therapies.
To ensure informed dialysis selection, participants are assessed for their confidence in dialysis decision-making and selection of dialysis modality at the end of the CPE session. Intent-to-Teach is assessed by confidence for dialysis decision making (defined by confidence rating of “quite confident” or “very confident”); those with suboptimal scores (“not at all confident” or “a little confident”) or “uncertain of the dialysis modality choice,” are advised to undergo repeat CPE sessions at an average of weekly intervals for a total of up to three counseling sessions (Figure 3). Our pilot studies show a vast majority of CPE recipients reach an informed decision by 3 sessions; when optional, 84% prefer to attend only one session, and when mandated for clinical care or research 96-99% of the patients reach informed dialysis selection by 3 sessions.[11, 21] Considering our preliminary data and to ensure the model is ready-for-dissemination, TEACH-VET allows CPE participants to pragmatically choose the method for CPE, either face-to-face, or through tele-medicine to the affiliated outpatient clinic or within their homes. We will analyze the differences in outcomes between these delivery methods in our secondary analyses.
Control/EUC arm: Participants in the EUC arm are provided printed hand-outs directing them to online self-learning CKD resources, freely available through several professional renal organizations, including the VA.[24, 25, 4, 5] While the investigators acknowledge the scientific need for an unaltered control arm, enhancing ‘usual care’ through provision of the self-learning resources was considered the appropriate ethical compromise. To mirror the expected duration between the pre-, and post-CPE data collection in CPE arm, EUC arm participants provide data for post-EUC knowledge, confidence in dialysis decision-making, and dialysis modality selection 10-days after the provision of the self-learning resources.
Qualitative study: The qualitative study employs a maximum variation sampling strategy to ensure a diversity of demographic and clinical characteristics.[27] Fifteen Veterans from each of the face-to-face-CPE, tele-CPE and EUC groups respectively are interviewed by telephone for 45-60 minutes using a semi-structured interview guide based on the Theoretical Domains Framework (TDF).[28, 29] Furthermore, an additional 15 Veterans who did not ultimately use their preferred dialysis modality are interviewed 90-days post-ESKD to explore experiences and barriers. The TDF supplies the working analytical framework for identifying factors that influence Veterans’ informed dialysis decision-making and experience with different dialysis modalities, including any perceived factors influencing dialysis decision-making, perceived barriers to home dialysis selection and use, and [for CPE arm] satisfaction with education session and counseling. The verbatim transcriptions for the audio-recorded interviews will be analyzed by two independent coders, organizing the data by domains of the framework, e.g. Knowledge: participants’ knowledge regarding dialysis; Beliefs about capabilities: participants level of confidence; Intentions: CKD management preferences; Social influences: influence of family members, friends, or caregivers; Beliefs about consequences: expectations about CKD management and evaluation of results; Optimism motivation to recommend dialysis to other patients; and Emotions: feelings about CKD treatment options.
Post-CPE/EUC follow up: Nephrology, and if not available, the primary providers for the participants are informed of the participants’ preferences for dialysis modality. This communique further instructs the providers regarding the need and importance of the pre-ESKD nephrology care, and the processes and desired timings for the peritoneal dialysis catheter insertion/vascular access creation. The providers are also informed about the contact information and approval processes for the VA ESKD services. Participants are then followed by EHR reviews at quarterly and by telephonic interviews at semi-annual intervals to assess their need/status of dialysis therapies and any changes in their preferred dialysis modality. Participants in CPE arm are allowed to re-access the audio-visual group education session independently throughout the study period. All outcome measures and their collections timings are available in Table 1.
Statistical Considerations: We used G*Power version 3.1.9.2 for sample size calculations, which are based on the primary outcome of (home) dialysis use. Using one-tailed test with alpha of 0.05 and 80% power, to detect doubling of home dialysis use in CPE relative to EUC arm —estimating home dialysis actual use to be 10% for EUC (based on the prevalent data) and 20% for CPE—yielded a total sample size of N=108 (54 per arm). Allowing the potential of attrition and missing data that cannot be accommodated by the proposed missing data handling techniques (up to 20% data loss), we will need 136 (68 per arm) to reach ESKD and use dialysis to allow detection of this clinically meaningful effect size. Considering we expect about 25% of the study participants to reach ESKD through the study period, we plan to enroll 544 Veterans with advanced CKD for the study.
Analytic plan:
We will use multiple regression analysis to examine the effect of the CPE intervention on Veterans’ knowledge of CKD and confidence in dialysis decision making post- intervention or EUC. We will include the baseline knowledge and confidence scores as covariates in the model, to account for pre- intervention/EUC values. We will use orthogonal Helmert contrast codes to test for the effect of both CPE as a whole (collapsed across telehealth and face-to-face delivery methods) vs EUC, and for the effect of tele-CPE vs face-to-face CPE. (Although we do not predict an effect of treatment delivery method, we have planned to include the comparisons derived by the Helmert contrast coding to test and account for any variance that may be introduced by different treatment delivery methods, should such variance/effect emerge.) We will use multiple logistic regression to examine the effect of CPE on Veterans’ initial selection of home dialysis; specification of this logistic regression model for home dialysis initial selection mirrors the regression models for confidence and knowledge, with the exception that the outcome is binary. Additionally, we will use logistic regression with Helmert contrast coding to compare home dialysis actual use between CPE and EUC groups, as well as between tele-CPE and face-to-face-CPE groups (within the overall CPE group). This multiple logistic regression for actual use of home dialysis constitutes the analysis for the primary outcome of this study.
For continuous secondary outcomes post-ESKD (e.g. HRQoL), we will use multiple regression analysis with Helmert contrast coding for CPE and EUC comparisons (as used in above regression models), an effect of dialysis modality actually used (home dialysis vs in-center dialysis), and interaction effects between the contrast codes and dialysis modality ([CPE-vs-EUC*Modality] and [tele-CPE vs face-to-face CPE*Modality]. Where applicable, we will include the outcome’s baseline scores and/or other relevant covariates. For dichotomous secondary outcomes post-ESKD (e.g., inpatient initiation of dialysis), we will use multiple logistic regression analysis, with the specification of this model mirroring that for continuous secondary outcomes post-ESKD, with the exception that the outcome is binary. Finally, for the secondary outcome of time to ESKD, we will calculate a Kaplan-Meier estimate.
For qualitative sub-study, TDF will supply the working analytical framework.[28] Two researchers will independently code first few transcripts using the framework, reading transcripts line-by-line to capture as many behaviors, values, emotions, and impressions as possible, and comparing results to ensure everything relevant was coded according to the constructs of the framework. An iterative process will be used to refine themes from the framework based on patterns in the data, generating a thematic map.[30] This will provide in-depth understanding of the barriers Veterans’ experience in acquiring the knowledge needed to manage CKD, and facilitators involved in their selection and use of a post-ESKD management strategy.