Efficacy of Educational Interventions in Improving Measures of Living-donor Kidney Transplantation Activity: A Systematic Review and Meta-analysis

Abstract

Background. 

To address patient-level barriers to living-donor kidney transplantation (LDKT), centers have implemented educational interventions. Recently, some have highlighted several gaps in knowledge and lack of evidence of efficacy of these interventions. No review has synthesized the available data.

Methods. 

We conducted a systematic review and meta-analysis of studies conducted to increase measures of LDKT. Outcomes of interest were LDKT rates, donor evaluation, donor contact/inquiry, total transplantation rates, and change in knowledge scores and pursuit behaviors. A literature search was conducted across 7 databases from inception until 2017. Educational interventions were a decision/teaching aid alone or with personalized sessions. Comparator was another intervention or nonspecific education. Random effects meta-analysis was performed to pool risk ratios (RRs) across studies.

Results. 

Of the 1813 references, 15 met the inclusion criteria; 9 were randomized control trials. When compared with nonspecific education, interventions increased LDKT rates (RR = 2.54; 95% confidence interval [CI], 1.49-4.35), donor evaluation (RR = 3.82; 95% CI, 1.91-7.64), and donor inquiry/contact (RR = 2.41; 95% CI, 1.53-3.80), but not total transplants (RR = 1.24; 95% CI, 0.96-1.61). Significant increased mean knowledge scores postintervention was noted, and most showed favorable trends in pursuit behaviors. Quality across the studies was mixed and sometimes difficult to assess. The biggest limitations were small sample size, selection bias, and short follow-ups.

Conclusions. 

Educational interventions improve measures of LDKT activity; however, current literature is heterogeneous and at risk of selection bias. Prospective studies with diverse patient populations, longer follow-ups, and robust outcomes are needed to inform clinical practice.

INTRODUCTION

Living-donor kidney transplant (LDKT) recipients when compared with deceased-donor recipients have superior patient and graft survival.¹ LDKT has the potential to provide early access to transplantation. In addition, LDKT recipients experience lower short-term rejection rates and have an improved quality of life.^2-4 Despite these benefits, LDKT rates at many transplant centers are low and have stagnated across North America.¹

Several barriers in terms of access to LDKT are well recognized. Two major patient-level barriers are recipient’s discomfort to discuss issues related to donation with potential donors and lack of knowledge about LDKT.^5-13 Hence, some transplant and dialysis centers have implemented educational interventions to address them, for example, deploying a pamphlet or conducting a teaching session. Recently, a scoping review and a narrative review summarized these interventions and highlighted several gaps in knowledge and lack of evidence of efficacy.^14,15 Deploying resources to interventions that have minimal or no effect was questioned; although, interventions conducted at the home of recipients and in the presence of their social networks were thought to be promising.¹⁴ Home interventions are more personal, interactive, and direct and have the potential to reach a much larger pool of potential donors.^{10,12,14,16-18} However, conducting and sustaining this intervention long-term might be challenging as it is time consuming, labor intensive, and costly.

Given this, the aim of our analysis was to systematically review studies that implemented educational interventions to increase measures of LDKT activity. In particular, we wanted to analyze if these interventions were effective in improving LDKT rates, donor evaluations, donor contact/inquiry, total transplants (living and deceased donor), knowledge scores, and pursuit behaviors.

MATERIALS AND METHODS

Definitions

An educational intervention was defined as any tool implemented to increase living donation. An intervention could be, but not limited to, a brochure, pamphlet, video, website, teaching session, or smartphone application. It was designed specifically to educate participants on living donation. Thus, an education intervention could be decision/teaching aids (DTAs) that are evidence-based tools designed to help participants in making specific choices.¹⁵ Interventions could also be DTA with personalized sessions (DTAP). This was further subdivided into small group sessions (SG) where multiple recipients with their social networks met at the transplant center, and 1:1 where an intervention was directed exclusively at one recipient and/or their social network either at home or at the transplant center. To summarize, an intervention could be a DTA alone, DTAP-SG, or DTAP-1:1. Comparator could be another intervention or nonspecific education (NSE). Social networks were defined as friends and family members of the recipient.

Outcomes

Outcomes of interest were as follows: number of LDKT, number of recipients who had at least one donor who started the LDKT evaluation, number of recipients who had at least one donor contact at the transplant center to inquire about donation, total number of transplants (living- and deceased-donor kidney transplants), and mean change in knowledge scores pre- and postintervention. We included total transplants, as some have described increased rates of deceased-donor kidney transplants with these educational interventions.¹³ Education interventions could improve recipient awareness of the benefits of transplantation in general, which could lead to increased overall transplantation, independent of LDKT. Thus, we wanted to capture the total number of transplants. Finally, we also performed a descriptive analysis of LDKT pursuit behaviors, such as comfort, willingness, and the attitude postintervention.

Study Selection

We excluded studies pertaining to the general public, policy, or center-level changes, that is, participating in a paired kidney exchange program or conducting desensitization. We excluded studies where the participants were any party other than the transplant recipient, living kidney donor candidate, and the social network. We also excluded studies specifically pertaining to increasing registration as an organ donor or whose primary goal was to increase deceased donation. Studies were excluded if there was no comparator group. Finally, published abstracts were only included if the protocol and study design had previously been published outlining the details of the intervention and the abstract summarized results in detail to allow for data extraction.

Search Strategy

The literature search was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.¹⁹ The review was registered in Prospero (ID: CRD42018053370). Relevant studies for this review were obtained from 7 databases until November 2017. Grey literature was verified for research in progress (including clinicaltrials.gov) and conference abstracts. The search strategy combined the concepts of “Kidney Transplantation,” “Living Donors,” and “Patient Education,” found as both text words in the Title/Abstract/Keywords, and in the Subject Headings (Table S1, SDC, https://links.lww.com/TP/B725). When duplicate reporting of the same study or patient group was noted, the more updated publication was selected and duplicates were discarded. The reference articles from each identified study were reviewed to identify any additional relevant studies; however, no other reference was found.

Data Extraction and Quality Assessment

Study eligibility was individually determined independently by 2 of the study authors (S.S. and A.A.). The studies selected for full-text analysis were then independently screened by the same 2 authors. Data extraction (Table S2, SDC, https://links.lww.com/TP/B725) was performed by 1 author (S.S.) and then verified by another (A.A). Whenever needed, we contacted the authors of the selected articles to assist with data collection. They were contacted at least twice, 1 month apart, before exclusion. Any discrepancies were resolved by discussion between both authors till a consensus was reached. Assessment of bias for randomized and controlled studies was done using the Cochrane Collaboration’s tool for assessing risk of bias.²⁰ Assessment of the quality of studies in observational studies was not done, given the inherent high risk of bias and lack of key information to allow bias assessment.

Data Synthesis and Analysis

For the outcomes of LDKT, LDKT evaluation, LDKT inquiry/contact, and total transplants, we estimated the risk ratio (RR) comparing the intervention and comparator groups for each study. For studies with zero outcomes in either group, we added 0.5 to the outcomes. For the outcome of knowledge, the mean differences between the intervention and comparator groups were calculated. Results were summarized in forest plots. To obtain quantitative summaries of the impact of educational interventions, we performed meta-analyses using random-effects models applying the “inverse-variance” method for weighting individual studies.²¹ It should be noted that though the forest plots included all studies, meta-analysis models only included those studies that used NSE as the comparator. Heterogeneity was examined using Cochran Q and I² statistics. The I² statistic ranges from 0% to 100%, with lower values taken to indicate lesser heterogeneity. When the P value of the Cochran Q test was <0.05, there was a statistically significant evidence of heterogeneity. Statistical analyses were implemented using the metafor package within the R Statistical software package.²¹ We planned a subgroup analysis comparing DTAP-1:1 at home with all other interventions for the outcome of LDKT. In addition, we planned meta-regression models to look at the impact of 2 covariates on the outcome of LDKT: the presence of social network and the presence of previous donors as a part of the educational intervention. However, due to a lack of an adequate number of studies, these analyses could not be performed.

RESULTS

Literature Search

The initial search found 3863 records. Once duplicates were removed within EndNote, there were 1813 references to review. We then reviewed titles and abstracts and of the remaining records, we identified 78 studies potentially eligible for full-text analysis (Figure 1).

FIGURE 1.: Prisma flow diagram.

Studies Included

Full-text analysis led to 25 studies being excluded as they were out of scope to the study question. Another 19 studies were excluded as they were either conference abstracts or papers with preliminary results and follow up full-length articles or abstracts with more details were found. We were unable to perform data extraction in 6 studies.^22-27 In one study, the intervention strategy was unclear,²⁸ and in 5 studies there were no comparator groups.^29-33 Seven abstracts were excluded as full-length publications were not found. One conference abstract was included, as a previous publication had detailed the study details and abstract itself provided enough data for extraction.^34,35 Ultimately, 15 studies were eligible for analysis, of which 9 were randomized control trials^{9,13,16,17,35-45} (Figure 1).

Interventions

We found the following interventions: 4 were DTA-only,^37,40,41,45 3 DTAP-1:1,^16,39,44 5 DTAP-SG,^{13,36,38,42,43} and 3 had multiple arms.^9,17,35 In some studies, the comparator arm was also an intervention.^17,43,44 Details are summarized in Tables 1 and 2. Eight interventions included an input from previous donors, either in the form of a video or direct interaction with participants.^{13,35-38,40,42,45} Only some entailed coaching participants on how to approach donors.^13,35,36 Nine included the recipient’s social network.^{9,13,16,17,37-41,44} Overall, quality across the studies was mixed and sometimes difficult to assess given key missing information within the reporting of the studies (Table 3).

TABLE 1.: Demographic and study level data of analyzed studies

TABLE 2.: Intervention and outcomes of analyzed studies

TABLE 3.: Assessment of bias for randomized controlled trials only

Outcomes

LDKT

Eight studies reported LDKT rates.^{13,16,17,38,39,41,44,45} One study with a crossover design could not be included in the final analysis.¹⁶ As demonstrated in Figure 2, overall, most educational interventions tended to favor LDKT. When pooling results from studies where the comparator group was NSE, the RR of LDKT for the intervention group versus NSE was 2.54 and 95% confidence interval (CI), 1.49-4.35. There was no evidence of heterogeneity across studies (I² = 0%, P for heterogeneity = 0.68). There was no enough data to conduct a subgroup analysis comparing DTAP-1:1 at home with other interventions. However, there were 3 studies that compared DTAP-1:1 at home intervention with NSE, DTA, and DTAP-SG, and the RR for LDKT were 4.00 (95% CI, 1.41-11.36), 1.72 (95% CI, 1.12-2.64), and 1.81 (95% CI, 0.58-5.65), respectively.

FIGURE 2.: Education interventions and living-donor kidney transplantation. We estimated the RR comparing intervention and comparator groups for each study. For studies with zero outcomes in either group, we added 0.5 to the outcomes. We included all available studies for a given outcome in a forest plot. However, the meta-analyses for each outcome were based only on those studies with nonspecific intervention as the control group. CI, confidence interval; DTA, decision/teaching aids; DTAP, DTA and personnel – (1:1 vs small group [SG]); LDKT, living-donor kidney transplantation; NSE, nonspecific education; RCT, randomized and controlled trial; RR, risk ratio; SG, small group sessions.

Donor Evaluation and Contact/Inquiry

Four studies reported on donor evaluation^17,38,39,44 and five studies reported on donor contact/inquiry.^{17,38,39,42,44} Educational interventions favored both outcomes of donors initiating an evaluation and contacting the transplant center inquiring about donation. When compared with NSE, the RR of donor evaluation in the intervention group versus NSE was 3.82 (95% CI, 1.91-7.64) and donor contact/inquiry in the intervention group versus NSE was 2.41 (95% CI, 1.53-3.80). There was no evidence of heterogeneity across studies (I² = 0% for both outcomes, P for heterogeneity = 0.84 and 0.76, respectively; Figure 3).

FIGURE 3.: Education interventions and donor evaluation, donor contact/inquiry, and total transplants. We estimated the RR comparing intervention and comparator groups for each study. For studies with zero outcomes in either group, we added 0.5 to the outcomes. We included all available studies for a given outcome in a forest plot. However, the meta-analyses for each outcome were based only on those studies with nonspecific intervention as the control group. CI, confidence interval; DTA, decision/teaching aids; DTAP, DTA and personnel – (1:1 vs small group [SG]); LDKT, living-donor kidney transplantation; NSE, nonspecific education; RCT, randomized and controlled trial; RR, risk ratio; SG, small group sessions.

Total Transplants

Five studies reported on deceased-donor kidney transplant rates and from which total number of transplants were ascertained.^{9,17,38,39,44} Education interventions targeting LDKT did not lead to higher total transplants, when compared with NSE (RR = 1.24; 95% CI, 0.96-1.61). There was no evidence of heterogeneity across studies (I² = 0%, P for heterogeneity = 0.73; Figure 3).

LDKT Knowledge

Change in knowledge scores was captured by 10 of the 15 studies^{16,17,35-40,44,45} and in 6 was the primary or 1 of the primary outcomes of interest (Table 2). No uniform scale was used to measure knowledge scores. Two used the Rotterdam Renal Replacement Knowledge Test, a 21-item self-report questionnaire.^16,39 The remaining studies used investigator-developed questionnaires, and questions were either true or false statements and multiple-choice questions; number of questions ranged from 9 to 62. Meta-analysis was not performed given lack of a standardized questionnaire. In addition, data could not be synthesized in 3 studies as SDs were not provided.^17,35,45 However, as demonstrated in Figure 4, overall educational interventions were associated with a significant increase in standardized mean difference in the knowledge scores before and after the intervention when compared with the control group (Figure 4).

FIGURE 4.: Education interventions and mean change in knowledge scores before and after the intervention. The mean differences in knowledge scores were calculated. No uniform scale was used to measure knowledge scores, and data could not be synthesized in 3 studies as SDs were not provided. Meta-analysis was not performed given lack of standardized questionnaires. CI, confidence interval; DTA, decision/teaching aids; DTAP, DTA and personnel – (1:1 vs small group [SG]); MD, mean difference; NSE, nonspecific education; RCT, randomized and controlled trial; SG, small group sessions.

LDKT Pursuit Behaviors

This was the most prominent outcome studied: 12 studies analyzed several different LDKT pursuit behaviors (Table 2).^{9,16,17,35,36,38-40,42-44} These included pursuit behaviors such as considering LDKT, endorsing LDKT, readiness to pursue LDKT, attitude toward LDKT, self-efficacy, willingness to pursue or accept LDKT, discussion with potential donors, frequency of communication with potential donors, and comfort with pursuing LDKT. In 6 studies, it was the primary or one of the primary outcomes of interest.^{9,16,36,39,40,43} Analysis ranged from assessing behaviors using yes/no questions or grading answers on a scale.^{13,17,36,38,44} Some used robust theoretical models to analyze these behaviors such as the Transtheoretical Model and Stages of Change, and the Attitude-Social Influence-Efficacy model.^16,35,39,43 In 8 studies, some, if not all, behaviors improved favorably with various LDKT interventions.^{9,13,16,17,35,40,42,44} In 2 studies, outcomes were different based on the behavior assessed.^17,40 Four studies reported no statistically significant change in these behaviors.^36,38,39,43 Interestingly, in some studies, no significant changes were reported among recipients; however, when accounting for all attendees, significant changes in attitudes were noted.^13,16,17,39

DISCUSSION

This systematic review reports that educational interventions implemented to increase measures of LDKT activity were effective. When compared with NSE, an intervention targeting education related to living donation was associated with 2.5 higher probability of LDKT, 3.8 higher probability of donor evaluation, and 2.4 higher probability of donor contact. It was also noted that 67% of these studies reported favorable changes in LDKT pursuit behaviors and in most studies significantly higher change in knowledge scores was noted in the intervention group. To our knowledge, this is the first systematic review and meta-analysis evaluating the efficacy of educational interventions in improving LDKT rates and other measures of LDKT activity.

Most dialysis and transplant centers vary in how well they inform and educate patients about LDKT and doing so is often not the standard of care.⁴⁶ When done, it is nontailored and nonspecific to patients’ state of readiness, and culture and educational background.^18,46 Also, NSE can be ineffective due to time constraints, inability to reach the recipient’s social network and placing the onus of finding donors on the overwhelmed recipient.¹⁸ Hence, home interventions are thought to be very promising as trained professionals can tailor education to family dynamics and promote patient advocacy and shared decision-making.^14-18 However, this intervention is quite resource intensive and might not be sustainable at most dialysis and transplant centers.^14,44 We report that even small interventions, such as DTA, can increase measures of LDKT activity. DTAs are designed to supplement clinician counseling and have been shown to improve knowledge, communication, and health outcomes.¹⁵ Home-based interventions may be more efficacious; however, due to the lack of available studies we cannot conclusively demonstrate this. Overall, we report that any of these interventions is superior to NSE, and this should inform decision makers at the dialysis and transplant centers.

To inform centers on the design of an educational intervention, a few observations merit discussion. First, the presence of social networks can be very influential in the communication and decision-making surrounding LDKT.^13,18,47 Indeed, we report that most studies included the recipient’s social network. Second, potential donors and their recipients want to personally speak with previous donors and recipients.⁴⁸ We note that 53% of the studies included previous donors narrating their experiences with LDKT. Third, coaching participants how to discuss donation with potential donors is a major barrier to LDKT.^13,35,36 We note that only 20% of the studies included this strategy. Fourth, it has been suggested that even for medical personnel, shorter length of time would better suit an adult learner to ensure knowledge retention.⁴⁹ Indeed, most studies had multiple, interactive, and brief sessions.^{9,13,16,35,37-39,44} Finally, some literature has suggested that certain patient-level factors can be used to tailor specific interventions, such as, more willingness to discuss LDKT, stage of readiness, and larger social network.^18,47 During data extraction, we could not tease out these details to conduct such an analysis.

This systematic review is the first to summarize educational interventions designed to increase measures of LDKT activity. The 15 studies were published over 19 years and included studies from around the world and included patients with different cultural and ethnic backgrounds. We systematically assessed the quality of these studies and quantified the data. Though the educational interventions were somewhat different across studies, we reasoned that the contents were quite similar and many were developed using the International Patient Decision Aid Standards.^14,15 Therefore, we felt that it was reasonable to pool across studies. However, it should be noted that our pooled result was not representative of any one type of intervention but rather an aggregate across the different types of educational interventions. Some limitations of our approach deserve discussion, in particular, selection bias (Table 3). In some studies, participation was restricted by geography,^17,38 language consideration,^9,16,17,36 ability to use a computer,³⁷ and physician preference.⁹ Overall, quality across the studies was mixed and sometimes difficult to ascertain. Only some studies investigated protocol adherence and had a quality assurance framework in place.^9,37,39 For the outcome of knowledge, most studies used investigator-developed questionnaires, only a few of which had been externally validated (Figure 4). In addition, several authors acknowledged that spurious conclusions may have resulted from small study samples and shorter follow-ups^38,42,45; however, by integrating them together, we hoped to have addressed some of these limitations. The small number of studies may also have limited our ability to detect heterogeneity across studies.

Despite these limitations, the findings of this review and meta-analysis have implications to both clinical practice and public health policy as there is a considerable push for LDKT. Our results suggest that even simple interventions, such as DTA, are superior to NSE in improving measures of LDKT. We recommend that these efforts be made in conjunction with other center- or policy-level interventions that could interact with or enhance LDKT. We, however, recommend that future studies have an evaluative and quality improvement framework in place, have longer follow-ups, and use robust outcomes, such as LDKT and time to transplantation. This is because surrogate measures, such as donor contact, may not necessarily translate to LDKT. An in-center quality control project revealed that >50% of the donors who contacted our transplant center were refused at the first contact, and from the remaining about 20% were noncommitted and did not contact the center again. Also, we recommend that follow-up in studies be longer than 1 year as the median duration of donor evaluation is reported to be 10 months.^50,51 Finally, given the significant disparities in access to LDKT,^7,8,52 efforts should be made to decrease selection bias, and to incorporate wider interventions applicable to patients across different races, ethnicities and to those of different literacy and socioeconomic status.

In conclusion, this systematic review demonstrates that educational interventions are effective in improving measures of LDKT; however, our findings are limited by the heterogeneity of the data, and inherent biases and limitations of these studies. Prospective studies with a diverse patient population, longer follow-ups, and robust outcomes are needed to inform clinical practice; many are currently ongoing.^34,53-59 In the meantime, centers could consider simple, less expensive interventions, such as DTA, to educate and assist potential donors and recipients.